Hey there.

No personal updates this week, just a lot happening, and I wanted to get this out to you. Today we've got a court case that should worry anyone using AI to screen resumes, Apple trimming staff out of teams that didn't pan out, and a viral app that sends your texts by virtual pigeon because apparently we've decided speed was the problem all along.

Let's get into it.

🌐 News Shortlist

1. Bias, Bots and Boundaries: Practical Tips for Avoiding Liability in AI Recruiting.

Recap: An August 21 explainer from HR Daily Advisor lays out where things stand with Mobley v. Workday, the federal case where applicants allege Workday's AI screening tools discriminated by race, age, and disability. A judge already let major claims move forward and certified a nationwide collective action under the Age Discrimination in Employment Act. The piece spells out what employers now need to do: inventory every AI tool touching recruitment, vet vendors before rollout, keep humans reviewing decisions, and document outcomes for bias.

I've been saying this since before it was a compliance headline. We have a human interviewing every single candidate we present. It's honestly a little wild that I have to keep mentioning that, like it's a differentiator instead of the baseline, but here we are.

Companies bought AI screening tools because they promised to remove a headache. What they actually did was remove a witness. When there's no human in the loop, nobody can explain why a candidate got filtered out, and now a judge is saying that's exactly the problem. The tools didn't eliminate the employer's responsibility. They just made it harder to see where things went wrong until a lawsuit forces you to look.

I use AI in my own process constantly. Every call gets an independent AI score, and it's a great sanity check. What it doesn't get is the final call. A recruiter still has to sit across from that person, ask follow-up questions, catch what a resume parser never will. Take that out and you're not innovating, you're just outsourcing your legal exposure to a vendor.

Advice: If a tool is making pass or fail decisions on your candidates without a person reviewing them, that's your biggest open risk right now, not your biggest efficiency win. Fix that before your legal team finds out about it from a subpoena.

2. Apple cuts jobs in Siri, Vision Pro immersive video and gaming teams.

Recap: Bloomberg reported on August 21 that Apple cut more than 200 roles across its Siri, Vision Pro immersive video, and gaming software teams, citing underperforming product bets. Apple said it would create new roles as part of the change even as it trims existing ones.

This is one of the biggest, best-resourced companies on the planet, and they still got the bet wrong on where to put headcount. That should tell you something. Money and scale don't fix a hiring decision made on hope instead of judgment.

What gets me is the pattern. A team gets built out fast around a product that's supposed to be the next big thing, and eighteen months later half of it gets cut because the thing didn't land the way anyone expected. I see smaller versions of this constantly with founders who hire five people quickly for a new initiative instead of one really strong person first. You end up paying for the headcount either way, you just pay for it twice, once in salary and once in severance.

Precision beats volume even when you're Apple. We tell clients this constantly: it's better to bring on one excellent person and see if the bet works before you scale the team around it. Nobody gets credit for how fast they hired if the whole team gets unwound a year later.

Advice: Before you staff up a new initiative, hire the smallest team that can prove the idea works. Scale the headcount after the bet pays off, not before.

3. Carrier Pidge messaging app says "waiting is the point."

Recap: Carrier Pidge, built by one developer, Noah Iarrobino, in about three weeks, sends your texts via a simulated carrier pigeon that flies at 110 mph. A message from LA to New York takes roughly 22 hours. There's a small chance, 0.2 percent, that your pigeon just dies en route. The app went viral starting August 12 and picked up mainstream coverage through August 25.

Okay, I need everyone to sit with this for a second. We spent twenty years building infrastructure to get a message from one phone to another in under a second, and the breakout hit of the summer is an app that makes it take a day and sometimes kills the messenger. I love it. I would not have built it, but I love it.

Here's the thing nobody wants to admit: this guy isn't really selling a messaging app. He's selling nostalgia for a world where you had to wait for something and that was normal, and he found a way to package that feeling and put it back on your phone. Everyone else in messaging spent a decade racing to zero latency. He looked at that whole industry and sold people the exact opposite of what it optimized for. Three weeks of vibe coding and now I'm talking about a pigeon on a Tuesday. That's the kind of insight that makes me want to buy him a drink and also feel a little bad about myself.

Also, and I say this as someone who has read a few too many term sheets, the monetization is aviaries, plural, so you can send more than one pigeon at once. I don't know what an aviary costs. I don't want to know. I just respect that somebody figured out how to charge you extra for owning more birds you cannot see.

Advice: If your product roadmap this quarter is "make it faster," go read about the pigeon app and ask yourself what the opposite bet looks like. Sometimes the whitespace is just doing the thing everyone agrees is dumb, on purpose, well.

That is it for this week. What ties the first two stories together is judgment, whether it's a court reminding you that a human still has to own the call an algorithm makes, or Apple learning that headcount built on a hunch gets unwound the same way it got built. The pigeon story just reminded me that not every good idea has to be the smart one on paper. Sometimes it just has to be the one nobody else was weird enough to ship.

Until next time,

Joseph Burns
CEO & Founder, Lupa Hire

Hey there,

Quick recap before we get into this week's stories. The death toll from the earthquake in Colombia keeps climbing. People and organizations have donated, and some real money has started showing up for the rebuild, but the total cost of putting the country back together is going to be far more than what's come in so far. If you haven't given yet, Cruz Roja Colombiana is still taking donations directly, and their relief teams are still deployed across Chocó and Valle del Cauca. Thank you to everyone who already gave last week.

Argentina's economy is straining in a way that should matter to anyone hiring there. The July jobs report came in soft, and the real story was buried in the revisions nobody read past the headline. And a new survey finally put a number on something recruiters have suspected for a while: a lot more job seekers are using AI than they're letting on.

Let's get into it.

🌐 News Shortlist

1. Argentina's Economy Is Sending a Warning Most Employers Aren't Watching.

Recap: Argentina's country risk index climbed to its highest level in two months during the week of August 17-18, even as the government posted a primary fiscal surplus. Economists are bracing for a second-quarter GDP contraction when official figures publish, pointing to record consumer goods imports straining the currency and a widened subsidized electricity allowance adding pressure to public spending.

That makes a rising risk number next to a fiscal surplus a strange pairing. Investors are confused about the direction of the fix and betting against whether it sticks, and a weaker peso is the part of that story that lands on a client's desk first.

There's a mistake I keep watching companies make when a currency wobbles like this: they treat it as a discount. I've watched a company anchor an offer to the worst possible peso reading, save maybe a few hundred dollars a month, and then spend three months wondering why the hire didn't stick. A rough currency number isn't a coupon. It's a reason the person you're trying to hire has more offers to weigh than usual, not fewer. Pay someone what the local market says they're worth and you get someone excellent for a fraction of what they'd cost in the US. Pay them off a headline instead, and you get whoever's left standing when everyone better already said no.

Advice: If you've got people, contractors, or vendors in Argentina, check how they're paid and how exposed that setup is to peso swings. First time hiring there? Budget for the role, not the exchange rate. Then keep an eye on the jobs data, because the same lesson about reading past the first number shows up there too.

2. July Jobs Report Lands Soft, and the Revisions Are Worse Than the Headline.

Recap: The Bureau of Labor Statistics released the July jobs report on Friday, August 7. Nonfarm payrolls fell by 23,000, and the unemployment rate rose to 4.1%. The bigger signal was in the revisions: May's already-modest gain was cut by 66,000 to just 63,000, and June's number was cut by 37,000 to 20,000. Combined, May and June came in 103,000 jobs lower than first reported.

One soft month doesn't mean much on its own. Three revised down in a row does. Nobody talks about the revisions because they're boring, but they're the part that actually tells you something: the market was weaker in real time than the first numbers let on.

That gap between the headline and the revision is exactly where a bad hiring assumption hides. A lot of teams have a role sitting open right now because someone decided months ago that good candidates were scarce, widened the search, and settled for whatever showed up. If the market was actually softer than the July number first suggested, that assumption was wrong when it was made, and it's more wrong now. The fix: match quality over volume, the same bar as before, applied to a pool that's better than anyone assumed.

Advice: One weak month, ignore it. A run of them, go back and check whether the "we can't find anyone good" excuse still holds. It usually doesn't.

3. Job Seekers Are Quietly Using AI to Land More Offers, and Most Won't Tell You.

Recap: A new survey of US job seekers found that a little over a third used AI somewhere in their search, and that group landed noticeably more offers than the ones who didn't. A meaningful share let AI submit applications automatically, and a smaller group used it live during interviews. Most AI users said they'd never disclose that use to an employer.

Forget the disclosure number for a second. The offer gap is what should bother a hiring team. People using AI well aren't cheating; they're just faster at researching the role and showing up more prepared than the person who skipped it.

Nobody set a rule about disclosing it, so most people don't. That part's on us, not them.

What I hear from clients is that take-home tests are dead. Job posts get buried under applications an AI wrote and submitted before a person ever looked at the listing. Clients who've adjusted stopped asking whether someone used a tool and started asking a harder question in the room, live: can this person work through a problem in front of me, use AI well under pressure, and explain what they just did? You can't fake that on a Wednesday afternoon call.

Advice: Stop policing whether someone used AI. Move the real test into a live conversation where you can watch them think. Whatever got them in the door matters a lot less than whether they can actually do the job once they're in the room.

That is it for this week.

If there's one thread running through all of it, it's this: whether it's a shaky peso, a softer jobs report, or a resume that may or may not have been written by a machine, the temptation is to trust the number in front of you and skip the harder step of checking what's real. Spend the extra ten minutes finding out what's real before you build a decision on a headline. That is the whole job, honestly, and it's what we help clients do every week at lupahire.com.

Until next time,

Joseph Burns
CEO & Founder, Lupa Hire

Hey there,

My thoughts are with the people of Colombia. A magnitude 7.4 earthquake struck the department of Chocó on Monday morning, and rescue teams are still working to reach the hardest-hit areas.

I spent four years living in Bogotá, and when I think back, one of the things that amazed me was how quickly people helped each other when things went wrong. No one needed to organize it. If there was a blackout or a neighbor was having a tough time, someone would start cooking, and others would bring what they could. I've never seen people act faster to help than in Colombia.

This week, the situation is personal for us. Part of our team lives in Pereira, one of the cities most affected by the earthquake. They’re volunteering, giving out free lunches to families who lost their homes, and are buying medical supplies for hospitals that are already overwhelmed. I'll share two ways you can help below: one through an established organization and one directly to our Lupa team members in Pereira.

Two other things worth your attention this week. Washington just made it more expensive to keep an H-1B or L-1 worker on staff, not just to hire one. And a new report on enterprise AI agents shows a lot of companies handing out access nobody would approve if they stopped to think about it.

Let's get into it.

🌐 News Shortlist

1. A 7.4 Earthquake Hit Colombia. Here's How to Actually Help.

Recap: A magnitude 7.4 earthquake struck western Colombia on August 10, with an epicenter in San José del Palmar, in the department of Chocó, roughly 150 miles west of Bogotá. As of this week, more than 200 people are confirmed dead, over 1,200 are injured, and more than 3,700 remain unaccounted for. The quake was felt across much of the country and into Ecuador and Panama. Cali's mayor said 32 buildings have collapsed there alone, including a large part of a hospital. Colombia's Geological Service called it the strongest earthquake the country has recorded this century. The government has declared a national emergency, and rescue teams are still working through aftershocks to reach the epicenter itself.

I spent four years living in Bogotá, working alongside people there, and I've never met a country that executes harder under pressure. That's the same instinct on display right now, just at a scale no city plans for.

If you want my personal recommendation, there are two honest ways to help right now, and neither one is wrong.

The first is to go through an established institution. Cruz Roja Colombiana, the Colombian Red Cross, has activated its national emergency response and is taking donations directly through its campaign at ayuda.cruzrojacolombiana.org, or by transfer to its verified Davivienda account, 0560 4506 999 649, held under the Sociedad Nacional de la Cruz Roja Colombiana. That money moves through a trained relief network with people already deployed in Chocó, Valle del Cauca, and the other affected departments, and it's the safer choice if you'd rather not send money to an account you can't verify yourself.

The second is more personal, and it's also one of ours. Part of our team lives in Pereira and is volunteering with a restaurant in a relief operation for victims. Every peso sent their way goes straight to ingredients for more free lunches and supplies for hospitals that are stretched past their limit.

• Transfers within Colombia (Bre-B): 42096602

• From outside Colombia (ACH, United States): account holder Liliana Patricia Rodriguez Soto, account number 963815257949, routing number 043087080, SWIFT SSBAUS32, account type checking, bank SSB Bank, 8700 Perry Hwy, Pittsburgh, Pennsylvania 15237

Advice: If you have Colombian colleagues or candidates in your pipeline right now, give them room. Some of them are checking on family instead of checking their inbox, and that should never be held against them.

2. Washington Just Made It More Expensive to Keep an H-1B Worker, Not Just Hire One.

Recap: On August 10, DHS published a rule expanding the $4,000 H-1B and $4,500 L-1 "9/11 response" fee to cover extension-of-stay petitions, not only new hires and change-of-employer filings. It applies to employers with at least 50 US employees where more than half the US workforce holds H-1B or L-1 status. The fee dates to 2015 and originally funded biometric entry-exit programs, applying only when a visa was first granted or an employee switched companies. The rule takes effect 30 days after publication. It follows an appeals court decision earlier this month blocking the White House's separate proposal for a $100,000 H-1B application fee.

The $100,000 fee got the headlines and then got blocked. This one is smaller, already final, and hits a different moment in the process, which is why it matters more for planning.

Most companies budget for the cost of hiring someone on a visa. Far fewer budget for the cost of keeping that same person once the visa needs renewing. This rule changes that math for the employers most dependent on H-1B and L-1 talent, the ones where more than half the US workforce is already on one of those visas. Every renewal now carries a fee that didn't exist for that filing before, on top of whatever the next round of policy brings.

None of this touches how you build a team in Latin America. A senior engineer or operator you hire in Bogotá or Medellín never needs a visa, never has an extension date, and never becomes more expensive to keep just because a new rule got published. The companies leaning hardest on the visa route are the ones who will feel this compound the most.

Advice: If a meaningful share of your team runs on H-1B or L-1 status, run the true cost of the next three years of renewals against the cost of hiring the same role directly in Latin America. The visa route just got a recurring tax added to it. The nearshore route didn't.

3. 60 Percent of Enterprise AI Agents Have More Access Than They Need.

Recap: A new report from Opsin Labs found that AI agent use inside companies grew fourteenfold between January and June 2026. The average enterprise now runs roughly one agent, live or in draft, for every employee. Sixty percent of agents given access beyond their default settings were granted allow-all permissions rather than access scoped to their actual task, and 60 percent had capabilities that had grown past what they were originally built to do. Two out of three agents are built by employees without an engineering background, in sales, customer success, and operations, often faster than security teams can review what they're allowed to touch.

Weirdly, I have to say this about my own industry, but a human interviews every candidate before we present them to a client. I mention it because most of the software built to automate hiring skips exactly that step, and this report shows the same pattern happening everywhere else companies are handing work to agents.

The numbers describe a predictable failure mode. Someone in operations builds an agent to solve a real problem, gives it broad access because that's the fastest way to get it working, and nobody revisits that access once the agent proves useful. Multiply that by one agent per employee, and you have a company that granted itself a security problem in the name of moving fast.

The fix isn't to slow down on AI. It's the same discipline this newsletter argues for in hiring: keep a human checkpoint on anything that acts on your behalf, whether that's a recruiter's shortlist or an agent with access to your customer data. The companies that skip that step in hiring end up with bad placements. The companies that skip it with agents end up with a breach.

Advice: Before you approve an agent's access, ask whether a person would sign off on that same access if a stranger walked up and asked for it directly. If the answer is no, the agent shouldn't have it either.

That is it for this week.

Colombia is facing a real crisis, and the most useful thing you can do right now is donate. Washington just made the visa route to US talent more expensive to maintain, not only to start, which is one more reason to build directly in the region instead. And the AI agent numbers are a reminder that the same human-in-the-loop discipline we use in hiring belongs everywhere else you're putting these tools to work.

The trends that move first are the ones worth watching now. That's what this newsletter is here for. It's also what we help clients with every week at lupahire.com.

Until next time,

Joseph Burns
CEO & Founder, Lupa

Hey there,

Years ago I was a product manager at Rappi, right as Covid hit. I still remember the morning it broke. By 10 am, we already had more orders than we normally saw in a full day. Restaurants were shut, cities were locked down, and overnight a whole region decided it needed delivery. We grew eight times in a matter of days, building the plane while we flew it, hiring and shipping and patching things in real time.

What stuck with me was who pulled it off: a Latin American company, run by Latin American engineers and operators, absorbing a shock that would have flattened most teams. That changed how I saw the region.

So when I saw this week that Nubank's Mexican arm had become a licensed bank, I read it as more proof of what I watched happen at Rappi. It reinforced the same lesson: the people who build companies like that are the same people you can hire, and most US companies still aren't looking their way.

Two other things are worth your time this week: Colombia swears in a new president on Friday, in a ceremony that has thrown out a century of tradition. And the AI models took a real step up over the last few weeks, which changes how small a strong team can be.

Let's get into it.

🌐 News Shortlist

1. Nubank Just Became Mexico's Biggest Bank.

Recap: On July 10, Mexico's banking regulator authorized Nu México to operate as a full bank, the final step in a conversion from a limited license that began in April 2025. Nubank's Mexican arm already has more than 15 million customers, roughly 15 percent of the country's adults, making it the largest digital bank in Mexico. The license lets it lend and launch products like a payroll account. Nubank has committed $4.2 billion to Mexico through 2030, including $2.5 billion in capital spending over the next four years.

This headline marks a big moment for fintech. But what really matters for anyone building a team is the hiring lesson: Nubank grew by drawing on local talent across Latin America. Nubank began in Brazil and is now Mexico's largest digital bank, thanks to engineers, risk analysts, product managers, and operators from all over Latin America. MercadoLibre and Rappi followed a similar path. World-class companies are being built in the region by local talent, and those people are still here, now training the next generation.

I tell founders this all the time: when you hire in Latin America, aim for the best talent, not just the lowest cost. Many try to save 80 percent and end up with someone who needs constant supervision. If you save 60 percent and hire someone who actually built the product, the value is much greater. For example, a senior person who costs $100,000 a year here would cost nearly $500,000 in the US. That difference is the real opportunity.

It's important to be clear: Latin America is made up of many different markets. Mexico and Argentina are as different as Germany and France, and each country has its own strengths. I usually hire operations people from Colombia, look to Argentina for innovators, and go to Brazil for consultative sales talent. Nubank's success in Mexico shows that the talent pool in these markets keeps getting deeper. Companies should see the region as a place to build, because that is how you get the best people. Those who look for bargains get what they pay for.

Advice: Compare your offers in Latin America to what top local companies pay, not to outdated numbers. Hire for the role and the talent, not the discount. The best candidates know their value, so attract them with meaningful work and a reason to stay, not just a low salary.

2. Colombia Swears In de la Espriella Friday.

Recap: Abelardo de la Espriella is inaugurated as Colombia's president on Friday, August 7, in a ceremony that breaks a century of precedent by taking place outside Bogotá, at a university arena in Cali. Outgoing president Gustavo Petro is boycotting the event and refused to make a military base available for it; a traditional medal ceremony has been dropped, and the handover has played out against fraud claims and warnings about stability. De la Espriella ran on a pro-market program of tax cuts, deregulation, and looser labor rules, reversing the direction of Petro's proposed labor reforms.

We've been following this race since May, and now the chaos is out in the open. The new president is being sworn in far from the capital, the outgoing government is boycotting, and security is tight. That makes the transition look messy, and some of it truly is. But look past the drama, and the hiring signal is clear. Petro's labor policies raised costs with stricter contracts, limits on fixed-term hiring, and higher overtime expenses. De la Espriella is moving in the opposite direction. For US companies hiring for outsourcing, IT, or operations in Colombia, looser labor rules make it easier and cheaper to hire.

There's also a practical currency issue. With a new president and uncertainty around the peso, top senior talent in Colombia now prefers to be paid in US dollars. That isn't a problem. It's already how the best candidates think, and paying in dollars often seals the deal. Make this part of your offer from the beginning. More broadly, Colombia is one of the best values in the region right now. Bogotá became a true tech hub after Rappi's rise, and you can find skilled professionals in their late twenties or early thirties who are experts in their field and work independently. Employer-of-record and severance costs are also lower than in Mexico or Brazil, making total costs easier to manage. The uncertainty around Friday's transition will scare off some companies, but that hesitation is your chance.

Advice: If you've been waiting to hire in Colombia, the inauguration is your sign to act, not to hold back. Start reaching out now while the headlines keep others away, plan to pay senior hires in dollars, and move before a more business-friendly labor market attracts more competition.

3. The AI Tools Just Took a Leap.

Recap: Over the past few weeks, the major AI labs shipped a wave of stronger models. Anthropic released Claude Opus 5 on July 24, described as close to its top-tier intelligence at roughly half the previous price. OpenAI rolled out its GPT-5.6 family, and xAI released Grok 4.5, all within the same stretch. Together, these releases point to more capability at lower cost, which lowers the bar for what a small team can take on.

Whenever AI models improve, people say recruiting will become obsolete. The reality is the opposite. AI has made job boards less useful because anyone can send out thousands of applications, flooding every posting with noise. So companies need someone who can actually find and assess talent more than ever. These tools have made my work better, not unnecessary. That's how I see this new wave of AI: cheaper, more powerful models don't replace people. They make skilled people even more valuable. At our company, we use AI in every process, but a person still interviews every candidate before we present them to a client. It may sound obvious, but most automated hiring tools skip this step. Good judgment is rare. Building something with these tools is easy now. Building the right thing, in the right way, still requires a person.

Here's what this means in practice. People used to talk about the legendary 10x engineer. With today's tools, a strong senior person can be closer to 100x, since the AI handles the volume and the person provides direction. That widens the gap between a small, sharp team and a large one, making each great hire even more valuable than before. You can find this kind of talent in Latin America for a fraction of a US salary, where advanced tools and talent arbitrage come together.

Advice: Don't treat the latest AI model as your whole strategy. Focus on finding people who know how to use these tools well. Before you buy another subscription or platform, check if your team is making the most of what you already have. The real value comes from skilled people, not from having more software.

That is it for this week.

Nubank's new license shows that Latin America builds world-class companies with local talent, and you can hire that talent if you see the region as a place to build. Colombia's inauguration on Friday may look chaotic, but it signals a more business-friendly hiring environment, and the current uncertainty is a chance for companies ready to act. Plus, the latest AI tools make small, capable teams even more valuable.

The trends that move first are the ones to watch right now. That's the focus of this newsletter, and it's also what we help clients with every week at lupahire.com.

Until next time,

Joseph Burns
CEO & Founder, Lupa

Hey there,

Two countries I follow closely just changed governments within ten days. Peru welcomed Keiko Fujimori as president on Tuesday, and Colombia will see Abelardo de la Espriella take office on August 7. Both support pro-market policies, plan to cut taxes and relax labor rules, and are working to reassure investors that it’s safe to return.

If you’re building teams in the region, this shift matters more than the headlines suggest. When a government shows it wants investment, hiring conditions often start to change before new policies are in place.

There are two other things to watch this week. Big Tech companies reported earnings, and Wall Street pushed back on how much they’re spending on AI. The Fed kept interest rates steady, but three officials argued they should have raised them instead.

Let's get into it.

🌐 News Shortlist

1. Peru Swears In a Pro-Market President.

Recap: Keiko Fujimori was inaugurated as Peru’s president on July 28, after winning June’s runoff by fewer than 50,000 votes. She is the country’s ninth president in a decade. Fujimori, 51, leads Peru’s largest political party and was running for the fourth time. Her agenda centers on political stability, reducing crime, and pro-market economic growth. Her win, alongside Abelardo de la Espriella’s in Colombia, extends a run of market-friendly governments across the region.

Put the vote margin aside for a moment. The real story is bigger: Peru has had nine presidents in ten years.

This instability sets the stage for the election. Peru has changed presidents so often that businesses can’t plan around it, with impeachments, resignations, and even an arrest. Fujimori won by promising to end this cycle, but it’s still uncertain whether she can deliver. Winning by less than 50,000 votes and facing a divided congress is a tough position. Peru voted for stability, and now it needs to make that real.

When it comes to hiring, Peru is in a unique position. Its talent market is smaller than Colombia, Argentina, or Brazil, and US companies don’t pay much attention to it. That’s what makes it interesting. Lima has skilled engineers, finance experts, and operators. The time zone matches the US, and salaries are lower than in more popular markets. Its lower profile is the real opportunity for hiring.

I always tell clients that the real value in a market like this is finding senior talent. US companies would pay two or three times more for the same person at home, and that difference is the main opportunity. If you come to Peru to build something meaningful, the numbers add up. But if you’re only looking to save a little on junior salaries, it won’t work out.

The reality is that political risk in Peru is still there, and one election doesn’t erase ten years of it. A pro-market government can help by making things a bit more stable: attracting more foreign investment, bringing in more multinational employers, and increasing competition for talent over time. If Fujimori brings even some stability, today’s quiet hiring market will get busier.

The bigger trend is what matters: Argentina under Milei, Colombia under de la Espriella, and now Peru under Fujimori are all moving in the same direction. For US companies, a region that’s opening up to investment and easing labor rules is easier to build in and hire from. Early movers can set their standards before the competition arrives.

Advice: Add Peru to your list before everyone else does. The top talent in Lima isn’t applying for jobs, so you need to find and contact them directly. Start now, while most US companies aren’t looking there, and set your standards based on today’s market, not where it might go if stability takes hold.

2. Big Tech’s AI Spending Spooked Wall Street.

Recap: In second-quarter earnings reported over the past two weeks, Alphabet, Amazon, Meta, and Microsoft confirmed they are on track to spend roughly $725 billion combined on AI infrastructure in 2026, up about 77 percent from last year, with projections near $950 billion for 2027. Alphabet raised its 2026 capital expenditure guidance to as much as $205 billion and reported that its free cash flow turned negative in the second quarter for the first time since its 2004 initial public offering. Its shares fell more than 7 percent. Across the sector, investors sold off on concern that AI revenue is not yet growing fast enough to justify the spending.

For most of this year, the market rewarded any company that spent big on AI. That just changed, and Alphabet shows why.

Alphabet is the figure to focus on. After two decades of steady profits, the company went cash-flow negative for a quarter because of heavy spending on data centers. Investors used to see this as ambition, but now they see it as a risk, and the stock fell. The main question is simple: the money is being spent now, but the revenue to justify it is still mostly a promise. This ties back to a theme we’ve tracked all year. Much of the money for AI investments came from cutting jobs. Companies called it efficiency and used the savings for computing power, hoping it would pay off. Now, Wall Street is no longer taking that on faith and is asking the same question we’ve been asking about layoffs: where’s the return?

For founders, the key takeaway is not to mistake the AI spending race for your own strategy. Alphabet can spend $205 billion and absorb the loss. You can’t, and you don’t have to. Think about the scale difference: big companies can put 400 engineers on a problem, while you might have one. You won’t win by outspending them, but you can win with good judgment. With today’s tools, a strong senior engineer can be as productive as a whole team. You can hire that kind of talent in Latin America for much less than in the US. That’s a very different approach from what you hear on earnings calls.

Advice: Don’t compare your AI strategy to Big Tech’s spending. Instead, focus on whether your team can use the tools you already have. The real advantage comes from skilled people. Spending money you can’t afford won’t get you there.

3. The Fed Held, and Three Governors Wanted a Hike.

Recap: On July 29, the Federal Reserve held its benchmark rate at 3.50 to 3.75 percent in a 9-3 vote. Three regional Fed presidents, Beth Hammack of Cleveland, Neel Kashkari of Minneapolis, and Lorie Logan of Dallas, dissented in favor of a rate increase, citing inflation that has stayed above the Fed’s 2 percent target for more than five years. The Fed described economic activity as expanding at a solid pace, with continued uncertainty tied in part to conflict in the Middle East.

Rates stayed the same. What’s more interesting is the direction of the dissenting votes.

For most of the past year, the Fed has debated when to cut rates. This time, three officials wanted to raise them. That’s a big change and shows that those setting rates are more concerned about high inflation than a slowing economy. If you’re waiting for cheaper borrowing before expanding, it’s time to stop waiting. Lower rates aren’t coming soon, and they could even go up.

Higher rates for longer mean US labor budgets stay tight. Companies don’t add staff when borrowing is expensive, which is why hiring has been cautious even with a strong economy. This same pressure has made building teams in Latin America more appealing all year, and this Fed decision doesn’t change that.

Advice: Plan your hiring for the next year assuming rates won’t change. You can’t control the Fed, but you can control your costs. A senior hire who costs $300,000 or more in the US can be much more affordable in Latin America, and that difference is a reliable hiring advantage no matter what the Fed does.

That is it for this week.

Peru and Colombia are both moving toward more investment, creating opportunities for companies willing to enter these markets before they become crowded. Big Tech’s earnings showed that Wall Street is finally questioning the rush to spend on AI, echoing concerns about the layoffs that funded it. The Fed also made it clear that cheaper money isn’t coming, so US hiring budgets will stay tight and the case for hiring in Latin America remains strong. The main takeaway is the same: the shifts all point in one direction.

The trends that change first are the ones to watch right now. That’s the focus of this newsletter, and it’s also what we help clients with every week at lupahire.com.

Until next time,

Joseph Burns
CEO & Founder, Lupa

Hey there,

If you work with people from Argentina, you probably won't be able to reach them today. At 3pm Eastern, Argentina faces England in the World Cup semifinal in Atlanta, and the whole country will be watching.

I've been saying all tournament that the World Cup changes everyone's schedule. Today proves it more than ever, and this match means more than just football. The winner will play Spain in Sunday's final.

There are two more things to watch this week. The USMCA review I mentioned in April has officially begun and will affect cross-border hiring in Mexico for years to come. Also, the biggest tech companies have spent the past month buying startups, mainly for their teams.

Let's dive in.

🌐 News Shortlist

1. Why No One in Argentina Will Be Working Today.

Recap: Argentina plays England in a World Cup semifinal today in Atlanta. It's the first meeting between the two national teams in 21 years. The winner faces Spain, who beat France 2-0 to reach Sunday's final. Argentina got here past Switzerland in the quarterfinals; England came through its own bracket to set up the tie.

Most sports rivalries are just about the game. This one is different, and it's important to know that if you work with Argentines.

The two countries went to war in 1982 over the Malvinas Islands. It lasted ten weeks and cost close to a thousand lives, most of them Argentine. Four years later, the two teams met in the 1986 World Cup quarterfinal in Mexico City, with the war still fresh. Diego Maradona scored twice. The first was the "Hand of God," a goal he punched in with his fist and the referee allowed. The second, minutes later, was a run past half the England team that gets called the greatest goal ever scored. Argentina won and went on to lift the trophy, and for a lot of Argentines that afternoon settled something the war hadn't.

The story kept growing. In 1998, England's David Beckham was sent off against Argentina, and England lost on penalties. He was blamed for years. This is their first World Cup meeting since 2005. For Argentines, beating England is always about more than just the score.

So the practical reality for today is simple. If you have Argentines on your team, the workday is effectively over, and it isn't only them. Across Latin America, people with no stake in either country will still stop to watch this one because everyone knows what it means. Trying to run a normal Wednesday against that is a losing move.

Advice: Let your Argentine team have the afternoon off without making them ask. Move any urgent work to another day, and if you want to build goodwill, encourage them to watch the match before they even bring it up.

2. The USMCA Review Is Finally Here.

Recap: On July 1, the United States, Mexico, and Canada formally launched the first joint review of the USMCA, the trade agreement covering roughly $1.8 trillion in annual commerce. It's the first mandatory review built into a US trade deal. The Office of the US Trade Representative has said it will not renew the agreement in its current form, with rules of origin and the treatment of Chinese-made inputs as the central issues. Mexico's economy secretary, Marcelo Ebrard, said the treaty stays in force under an annual review process, and a third round of US-Mexico talks is set for the week of July 20.

I mentioned this back on April 1, when it was just a date to watch. Now it's happening, and it's the most important trade story for anyone building teams across the US-Mexico border.

People are worried about disruption, but that's not the case right now. Ebrard made it clear that the agreement remains in place and that there are no immediate changes to the rules, so nearshoring to Mexico still works this year. The real issue is a longer negotiation about North American trade through the 2030s, especially around China. The US wants to ensure Mexico isn't used as a backdoor for Chinese parts and investment, and this extends beyond just car factories.

For example, if you make or assemble products in Mexico using imported parts, the new content rules from this review will affect your costs and compliance for years. That's a supply-chain issue, but it's also a hiring one, because companies expand or freeze their Mexican teams based on how confident they feel about the rules.

Software, engineering, and operations talent hired remotely in Mexico isn't governed by rules of origin, so the day-to-day case for hiring there doesn't change with this review. The risk is second-order: if the review turns sour and investment sentiment cools, competition for talent softens, creating a window for companies that keep hiring while others wait to see how the politics land.

Advice: If your Mexican team works in software or operations, keep hiring, since this review doesn't affect remote talent. If you have a physical supply chain in Mexico, start reviewing your component sourcing now to prepare for stricter China-content rules ahead of the next talks.

3. Everyone's Buying Teams, Not Products.

Recap: The last month brought a run of AI acquisitions aimed at talent rather than revenue. Anthropic acquired Vercept, a startup building computer-use agents. Zendesk bought Forethought for AI customer support. Meta acquired a robotics-AI startup, and OpenAI continued a string of team-focused deals. SpaceX moved to acquire the coding startup Cursor in a deal reported at $60 billion. Meanwhile, Y Combinator's latest batch put roughly 196 startups in front of investors, overwhelmingly AI companies built by very small teams, with the fund now offering part of its investment in stablecoins.

The common thread in all these deals is that companies are buying people. Acqui-hires used to happen when a startup failed, and a big company picked up the engineers for less. That's not the case anymore.

Now, acqui-hires are a premium strategy. Companies are paying top dollar to bring in small teams who have built tough products, because a group that works well together and understands a problem deeply is the scarcest thing in the market. Vercept and Forethought weren't rescues. They were teams worth more inside a bigger company than outside it. Even the SpaceX and Cursor numbers, whatever the final figure, are a statement that the right group of builders is worth almost any price.

Y Combinator's latest group shows the other side. Nearly 200 companies, most focused on AI and most very small. The winning model now is a few strong people using AI to do what used to take fifty. That's what buyers are paying for, just earlier in the process.

The lesson for founders is to see what the market values: small teams of exceptional people who can build with these tools. You can build that team yourself, and you don't need a Silicon Valley budget. Senior engineers and operators like this are available across Latin America right now, for much less than an acqui-hire and without the hassle of merging teams.

Advice: Stop thinking about headcount and start thinking about density, meaning the smallest group of excellent people who can move quickly with AI. That's what the biggest companies are spending billions to get, and you can build it too if you hire for skill and judgment, not just numbers.

That is it for this week.

Today, Argentina pauses for a match that's about much more than football, and much of Latin America does too. The USMCA review has begun, and while changes won't happen overnight, it will shape investment and hiring in Mexico for years to come. The biggest tech companies have shown this month that a small, excellent team is the most valuable thing you can own, which is exactly what we help our clients build.

The trends that move first are the ones to watch. That's the purpose of this newsletter, and it's also what we help clients with every week at lupahire.com.

Until next time,

Joseph Burns
CEO & Founder, Lupa

Hey there,

I have to admit I was wrong this week. Two issues ago, I celebrated the USA winning its group, but on Monday, Belgium ended their run with a 4-1 win. So much for my predictions.

Argentina is now the only Latin American team left after an exciting comeback against Egypt. They were down 2-0 with about ten minutes to go and managed to score three goals. Colombia lost to Switzerland on penalties the same day. The final is on July 19, so there’s still more action ahead.

Before we move on, a quick update on Venezuela. Recovery from last month’s earthquakes has shifted to rebuilding. The death toll is now over 3,500, with more than 12,000 injured, and a UN-backed report estimates direct damage at nearly $37 billion. The fund I mentioned a few weeks ago is still open for donations, which might matter even more now that the world’s attention is shifting. Here’s the link to donate to Caritas.

There are two other stories worth your attention this week: Microsoft cut thousands of jobs while also funding a 6,000-person AI team, and Amazon has started using AI agents for part of its hiring process.

Let's get into it.

🌐 News Shortlist

1. Argentina Is the Last Latin American Team Standing.

Recap: The Round of 16 wrapped this week. The United States lost 4-1 to Belgium on July 6, ending the host nation's run. Colombia went out to Switzerland on penalties after a 0-0 draw. Argentina came back from 2-0 down to beat Egypt 3-2 and is now the only Latin American team left in the tournament. Argentina plays Switzerland in the quarterfinals on July 11 in Kansas City. The final is July 19 at MetLife Stadium.

This is important if you have Argentines on your team, but it actually matters for everyone. Football in Latin America is a regional affair. Millions of people will keep watching even if their own teams are out. Some might complain about Argentina, but they’ll still tune in.

It’s worth thinking about this for a moment, because it differs from how most US companies view the region. Many treat Latin America as if everyone follows the same calendar, but that’s not true. At the same time, it’s not just about national borders. Pride and passion cross borders in ways you can’t measure on a spreadsheet, and Argentina shows this both on and off the field. They have the highest agency in the region, are great problem-solvers, and are comfortable improvising when things go off track. That’s not by chance. It’s the same qualities that make Argentines strong hires for roles that need ownership and flexibility.

If you have employees anywhere in Latin America, expect productivity to drop until July 19. It’s a scheduling fact, just like how US offices slow down during March Madness. The quarterfinals and semifinals are on weekdays during working hours in the Eastern and Central time zones.

You can try to fight this, or you can plan for it. The teams that handle it best aren’t the ones insisting on business as usual. They move deadlines, avoid scheduling big interviews during matches, and let people watch without pretending nothing’s going on.

Advice: Accept the drop in productivity rather than fight it. Check the match schedule, avoid scheduling important deadlines or final interviews during those times, and let your Latin American team know it’s okay to watch. You’ll make up the hours after the 19th, and you’ll earn goodwill you can’t buy.

2. Microsoft Cut 4,800 Jobs and Funded a 6,000-Person AI Team the Same Week.

Recap: On July 2, Microsoft said it would cut about 4,800 jobs, roughly 2.1 percent of its workforce, with the reductions concentrated in its Xbox gaming division, and about 3,200 roles phased out through fiscal 2027. In the same period, the company committed $2.5 billion and 6,000 employees to a new unit focused on helping customers actually implement AI. Microsoft's chief people officer said the eliminated roles are not being replaced by AI, while acknowledging that AI is changing how the work gets done. The company continues to spend tens of billions on AI data centers.

Those two numbers together tell the story and fit the pattern we've tracked all year. Big Tech isn't so much deleting jobs as repricing and relocating them. Money and people are moving from older divisions into new roles focused on helping other companies use AI. That new implementation team exists because selling an AI model is easy, but getting a company to use it well is much harder. The gap between what AI can do and what companies actually use it for is where these new jobs are found.

For example, one of the fastest-growing roles is someone who works with a customer’s team to figure out why adoption has stalled. Microsoft’s people chief was careful to say the cut roles weren't replaced by software, and that's worth believing: this is a budget being reallocated, not a function being automated out of existence.

For founders, the takeaway is to watch where talented people are going. If Microsoft needs 6,000 people to help companies use AI, the most valuable skill now is a mix of technical know-how and the ability to help teams change how they work. That’s the kind of experienced, judgment-driven talent you can find across Latin America, and it’s more affordable to build a team like this now than it will be in a year.

I’ll add something from our own hiring data. PwC found that workers with AI skills earn 56 percent more than others in the same role. Upwork saw demand for AI skills on its platform jump 109 percent last year. That pay bump isn’t just for listing a chatbot on your resume. When we screen for AI skills, we ask what someone has built with it and what challenges they faced. If they only followed a tutorial, they’re not ready yet. If they solved a real problem, that’s what matters.

This is the same idea behind what I call the 100x recruiter. A great chess player alone loses to AI, and AI alone loses to a great chess player using AI. The real advantage comes when good judgment is already present, whether it’s a recruiter or one of the 6,000 people Microsoft is hiring.

Advice: Don’t see Microsoft’s layoffs as a reason to stop hiring. Instead, look at the 6,000-person team as a guide. The most valuable hire now is someone who can make AI tools work within a team, and you don’t need a Big Tech budget to build that capability.

3. Amazon Put AI Agents in Charge of Part of Its Hiring.

Recap: Amazon released an agentic AI hiring product this year and has moved roughly 200 of its own HR roles to AI agents. The changes come alongside about 16,000 corporate job cuts, following roughly 14,000 in October 2025, which together account for close to 9 percent of Amazon's corporate workforce in three months. The company has framed the shift as part of a broader move to run leaner with AI handling routine tasks.

I have some experience here. I spent two years building an AI recruiter for Lupa before shutting it down, because the candidates we wanted didn’t respond. Good people with options don’t do AI interviews. They self-select out the moment they smell a bot, and the ones who stick around and game the process are disproportionately the ones you don't want. Kill the human touch, and you don't just lose warmth. You change who applies.

There’s a real case for efficiency here, and I won’t deny it. Much of recruiting involves coordination and scheduling, and AI agents handle them well. The risk isn’t the technology itself. The problem arises when the parts of hiring that need a human touch are handed off to software, along with the routine tasks. Screening, first impressions, and how you deliver a “no” all affect how candidates feel about your company. These are often the first things to get automated because they’re easy to script.

Amazon can handle the reputational hit if things go wrong, because it’s Amazon. Most companies can’t. When a candidate has a cold or confusing experience with a small startup, they don't shrug it off as the price of scale. They tell people. And in the markets where we hire, senior talent relies more on reputation and referrals than on job boards.

One helpful way to approach this is to split your hiring process into two parts. That’s how we do it at Lupa. AI handles research, market mapping, call recordings, prep, and recap notes. A person takes care of every conversation, every relationship, and every part of selling the role. Let AI manage logistics, scheduling, and status updates: the things candidates want done quickly. Keep people in charge of judgment and communication. Amazon is focused on volume, but if you’re trying to hire specific people, you need a different approach. If you copy the volume strategy, you’ll lose the candidates you want most. Always protect the human touch at the end.

Advice: Use agents for the mechanical parts of hiring and keep people on the decisions and the communication. Before you automate a candidate touchpoint, ask whether a real person would be comfortable being on the other end. Don't take the human out of Human Resources.

That is it for this week.

Argentina is the region’s last hope in the quarterfinals, and your Latin American team will be watching, no matter which country they’re from. Plan for a dip in productivity through July 19. Microsoft’s story shows that AI is still mostly about shifting money and people, not replacing them, and that the real work is helping companies use these tools. Amazon’s example shows what can happen when you let agents handle hiring, and why companies looking for specific talent should be careful about what they automate.

The trends that change first are the ones to watch right now. That’s the purpose of this newsletter, and it’s also what we help clients with every week at lupahire.com.

Until next time,

Joseph Burns
CEO & Founder, Lupa

Hey there,

My thoughts are with the people of Venezuela. Two large earthquakes struck northern parts of the country on June 24, and rescue teams are still working through the rubble to reach people.

I’ve worked with Venezuelans for years. They’re some of the most capable people I know, and they’ve built careers under conditions most of us never have to think about. Seeing them face this is hard.

I reached out to friends there to ask what actually helps right now. They said money for rescue efforts and shelter for families who’ve lost their homes. They pointed me to Caritas, a trusted option. Donate if you can. I’ll point to it again in the first story below.

Two other things stood out this week. Three labor reports showed layoffs are low, but getting rehired is harder. New hiring data show that half of rejected candidates never hear back from anyone, which is hurting companies’ reputations.

Let’s get into it.

🌐 News Shortlist

1. Two Earthquakes Hit Venezuela. Here’s How to Help.

Recap: On June 24, two major earthquakes struck northern Venezuela seconds apart: a 7.2 followed by a 7.5, with the epicenter in the state of Yaracuy. The worst damage hit La Guaira and Caracas. As of this week, more than 1,700 people are confirmed dead, over 10k are injured, tens of thousands are still unaccounted for, and roughly 59k buildings have been damaged or destroyed. International urban search-and-rescue teams from many countries, including the United States, are on the ground. The US government has pledged $150 million in aid. Rescue crews are still pulling survivors from collapsed buildings a week on.

Venezuela is a place full of people I work with and trust, and right now they need help.

If you’ve read this newsletter for a while, you know I don’t treat Venezuela as an abstraction. I’ve written about the World Baseball Classic win, about Maduro’s arrest, about the professionals who kept building through years most countries never have to survive. Now those same people are dealing with two earthquakes on top of everything else.

The friends I spoke to were clear about what matters at this stage. Money moves faster than goods, and it reaches rescue and shelter efforts directly. Caritas Venezuela has been among the first to respond, rooted in the territory since 1997 and maintaining a permanent presence in communities across the country.

If you employ Venezuelans, give them time. Some are checking on family and can’t work as usual right now. Let them know it’s okay to take the time they need.

Advice:

If you can, donate to a verified fund now. If you have Venezuelans on your team, let them know work can wait.

2. The Layoffs Cooled. Getting Rehired Got Harder.

Recap: Three labor reports landed this week and pointed the same direction. ADP reported that private employers added just 98k jobs in June, below the roughly 118k economists expected, with education and health services accounting for nearly half of the gains. Challenger, Gray & Christmas reported June layoffs cooled to 45,849, down 53 percent from May, but AI was the most-cited reason for cuts for the fourth month running, at 14,029 in June and 101,743 so far in 2026. Technology led every industry with 139,156 cuts year to date, up 83 percent from the same period in 2025. The Chicago Fed’s real-time model forecasts June unemployment at 4.36 percent, up from 4.30 percent, driven not by more firing but by a drop in the rehiring rate for unemployed workers, which fell to 44.45 percent from 45.69 percent in May.

The headline most people will read is that layoffs cooled. That’s true, and it doesn’t tell you much on its own. The more useful read is what comes next.

Look at the three reports together, and the picture is consistent. Hiring is slowing, not collapsing, and the slowdown is concentrated. Strip out education and health services from the ADP number, and the rest of the economy added roughly 50,000 jobs across every other sector combined. When one sector carries that much of the load, the market feels tighter to a jobseeker than the headline suggests.

The catch is in the rehiring rate. Layoffs barely moved, but the share of unemployed people finding their way back into work dropped again. That’s why the Chicago Fed expects unemployment to tick up. It’s more of a hiring problem than a firing problem. If you have a job, you’re likely keeping it. If you’re looking for one, the door is narrower than the low layoff numbers suggest.

For employers, the job market is slow. Fewer candidates have multiple offers, so you can take your time to find the right person. But waiting too long for the “perfect” candidate can cost you, including in Latin America.

Advice:

Be clear about what you need in a role and act when you find the right candidate. Don’t delay decisions just because the market feels slow.

3. Half Your Rejected Candidates Never Hear From a Human.

Recap: A 2026 Enhancv survey found that 50.5 percent of job seekers were rejected at least once in the past year without a single word from a human, and 63.8 percent of that group believed a machine made the call. Only 9.7 percent said an employer had ever clearly told them AI was involved in a decision. Separate research found 70 percent of candidates who went through an AI interview weren’t told upfront that AI would evaluate them, 34 percent came away with a more negative view of the employer, and 38 percent had already withdrawn from a hiring process specifically because it used an AI interview.

I wrote about this on LinkedIn this week because a reader sent me a rejection email that contradicted itself twice in three sentences. It said they’d picked other candidates, then said the role was closing, then encouraged the person to keep applying for the role that no longer existed. The candidate had no idea what actually happened, and that confusion was the whole message.

The data says this isn’t a one-off. Half of jobseekers were turned down last year without ever hearing from a person, and most of them assumed a machine made the decision. For example, a candidate completes an AI screening, receives an automated “no,” and never learns whether a human ever saw the application. That experience is what they remember, and it’s what they tell other people.

Why does this matter? Every candidate you reject could be a customer, referral, or future hire. Many companies automate rejections without thinking. A clear, honest “no” earns respect. Confusing, automated messages can hurt your reputation, especially if you’re hiring a lot.

Rejection is unavoidable. Confusion and disrespect aren’t. If you’re going to tell someone no, the least you can do is make the message clear and make sure a person stands behind it.

Advice:

Check the rejection messages your system sends and read them as a candidate would. If a person isn’t reviewing decisions or the wording isn’t clear, you’re risking your reputation for the sake of speed.

That is it for this week.

Venezuela is facing a crisis, and the most useful thing right now is to donate to a verified fund and give Venezuelan colleagues space. Job market reports show layoffs are low, but finding new work is hard, so clear, patient hiring helps. The AI rejection data shows that treating candidates well shapes your reputation.

The trends that move first are the ones worth watching now. That’s what this newsletter is here for. It’s also what we help clients with every week at lupahire.com.

Until next time,

Joseph Burns
CEO & Founder, Lupa

Hey there,

Three weeks ago, I told you this week would be the last clear sign before Colombia’s new era, whichever it may be. Far right-wing, pro-Trump Abelardo de la Espriella won the runoff by less than a point, the closest presidential margin in Colombia’s modern history.

I’m not Colombian, and I don’t vote there, so this isn’t about my personal preference. What matters now is how things will change for anyone doing business in Colombia, and those changes begin immediately.

Other news this week: The US central bank (the Fed) decided not to cut interest rates, citing a still-strong job market. In tech, more than 185k people have lost their jobs this year, mostly because companies are using AI to automate work. But even with all these layoffs, there are still 275k open jobs that require AI skills, jobs that most of the laid-off workers don’t have the training to fill. I’ll tell you more below.

Let’s get into it.

🌐 News Shortlist

1. Colombia Swung Right by Less Than a Point.

Recap: Abelardo de la Espriella won Colombia’s June 21 runoff with 49.6 percent, just ahead of Iván Cepeda’s 48.7 percent, a difference of fewer than 250k votes, making it the closest result in modern history. De la Espriella, a right-wing outsider often compared to Argentina’s Javier Milei, promised to shrink the state by about 40 percent, cut corporate taxes, broaden the tax base, and reopen Colombia to foreign investment and hydrocarbon production. He takes office on August 7. Markets rallied, and US Secretary of State Marco Rubio called de la Espriella to signal closer economic and security ties.

For three years, Colombia’s story was one of caution. Investors held back, foreign direct investment stayed below 2022 levels, and competition for senior talent in Bogotá and Medellín lagged behind the talent available. That was the window we kept talking about.

Sunday’s result is starting to close that window. De la Espriella ran on a platform that markets like, and they responded as expected: stocks rose and the peso strengthened. When a country signals it’s open for investment, multinational employers act before the new government takes office. They hire based on the rules they expect, not just the ones in place now.

The new president, however, faces a big challenge on a narrow margin. Winning by less than a point is not a strong mandate, and de la Espriella’s proposals are bold but lack detail. His confrontational style adds uncertainty, and Colombia’s fiscal deficit is expected to grow regardless of who is in charge. So, while this is a real change in direction, it’s also a contested one.

For hiring, this is what you need to know: the quiet market is about to get busier. The engineers, product leaders, and operators you can reach at current rates, with little competition, won’t stay available once investment increases. You’ll see this shift in salary expectations before it appears in any report, so act before the market tightens.

Advice:

If you’ve been waiting to build in Colombia, waiting now comes at a cost. Start reaching out while there are still few competing offers, and compare salaries to where they’re headed, not where they were last year. Move now so hiring stays ahead of the shift.

2. 275k AI Available Jobs. The People Being Laid Off Can’t Fill Them.

Recap: By late June 2026, tech layoffs had passed 185k, with 56 percent mentioning AI or automation. Meanwhile, about 275k AI-specific jobs remain open. ManpowerGroup’s 2026 survey of over 39k employers found that AI skills are now the hardest to hire for, with demand outpacing supply by about 3.2-to-1. Openings for machine learning engineers are 59 percent higher than before the pandemic, while general software engineering roles are down 49 percent. The jobs being cut (like customer support and QA) aren’t the ones being created. A 2025 MIT study found that 95 percent of enterprise AI pilots showed no measurable financial return, and Forrester found only 16 percent of workers had high AI readiness in 2025, a number expected to reach just 25 percent by the end of 2026.

Layoffs and hiring are happening at the same time, but most coverage treats them as the same story. They’re not. One is about jobs disappearing, and the other is about jobs that no one is prepared to fill.

If you look at job postings, demand for machine learning engineers is much higher than before the pandemic, while demand for regular software engineers is much lower. A laid-off support rep or QA analyst can’t make that jump in just a few months. These skills take time to develop, and most employers aren’t investing in them. Forrester found that only 16 percent of workers were ready to work with AI last year, and that number will only reach about a quarter by the end of this year. Companies are letting go of people with valuable knowledge, but don’t have enough trained staff to manage the new tools.

For example, a company might replace its content team with an AI workflow, only to find that no one can prompt, check, or fix what the tool creates. The MIT study highlights this problem: 95 percent of enterprise AI pilots showed no measurable return. Efficiency was assumed, not proven.

This situation creates two groups: a large pool of skilled mid-career professionals who lost their jobs because of AI that hasn’t delivered, and a small, highly competitive group of people who can actually do AI-focused work. Both groups reward the same approach: reaching out directly and assessing real skills instead of waiting for the perfect résumé.

Advice:

Before posting another senior AI job into a market with a 3-to-1 shortage, consider the talent you already have and the experienced people who were just let go. Training a strong operator to work with these tools is often faster and less expensive than competing for a title everyone wants, so prioritize internal and recently displaced talent first.

3. The World Cup Knockouts Start, and the USA Won Its Group.

Recap: The United States beat Australia 2-0 in Seattle on June 19 and won Group D. This is the first time the USA has topped its World Cup group since 2010. The 2026 tournament features 48 teams, with the top two from each group and the eight best third-place teams moving to the Round of 32. The USA’s knockout match is set for around July 1. On June 10, we mentioned research estimating up to $30 billion in lost US productivity during the tournament’s 39 days.

I predicted the USA’s success early and got some pushback for it. Now, after two straight wins and a group title, the knockout rounds are here. This is also when the schedule starts to affect teams across the Americas.

The productivity estimate we discussed on June 10 was always just an approximation. We see now how the schedule lines up. Knockout matches for Brazil, Argentina, Mexico, and now Colombia happen during working hours in the Eastern and Central time zones. Even your most responsible employees from those teams will be harder to reach on match days. Decisions will take longer, and interviews planned around elimination games may be delayed.

Timing is still the main issue. Hiring cycles in Colombia usually take six to eight weeks. Since the tournament runs until July 19, any search that hasn’t started yet probably won’t finish until August, which will push onboarding into the fourth quarter. I made this point before the group stage, but now there’s even less time.

Advice:

Work with the knockout schedule instead of against it. Schedule interviews outside of match times that matter to your candidates. If you want to hire in Latin America this quarter, start your search this week, the July 19 final is closer than you think, and delays will only compress your hiring window.

That is it for this week.

The labor market remains strong, which is why the Fed stopped promising rate cuts, and why those 275k jobs matter more than the layoff numbers. Plus, the World Cup knockouts are about to make your Latin America schedule more complicated than you might expect, so hiring now requires faster action.

The trends that shift first are the ones to watch right now. That’s the purpose of this newsletter, and it’s also what we help clients with every week at lupahire.com.

Until next time,

Joseph Burns
CEO & Founder, Lupa

Hey there,

I kept telling people to watch out for the USA team, but no one believed me. Well, here it is: the US beat Paraguay 4-1 on June 12 at SoFi Stadium. That's the biggest World Cup win for the US since 1930, impressive whether or not you're following the tournament.

PwC also released a labor market study this week, analyzing over one billion job postings from 27 countries. The main takeaway: AI is dividing workers into two groups, with a 42 percent gap in salary growth between them. Learn more below.

Tech layoffs are averaging 1,115 people a day. A Gartner study of 350 companies found that those making the biggest cuts aren't performing any better than those making smaller cuts. Colombia's presidential runoff is this Sunday.

Let's get into it.

🌐 News Shortlist

1. The Labor Market Is Splitting in Two.

Recap: PwC's 2026 Global AI Jobs Barometer looked at over one billion job ads in 27 countries and found that AI is splitting the labor market into two tracks. "Professionalised" roles, where AI takes care of routine tasks and lets people focus on judgment, are growing twice as fast as "democratised" roles and have 42 percent faster salary growth. Jobs needing AI skills grew 69 percent year over year, compared to 9 percent for the overall market. The wage premium for AI skills is now 62 percent, up from 57 percent last year. Companies using AI the most have grown their workforces by 52 percent since 2018, while less AI-focused firms grew by 36 percent.

Most reports on AI and jobs say, "some jobs are created, some are lost." That's true, but it doesn't really tell you what it means for you.

PwC's breakdown is more specific. In "professionalised" jobs like radiologists or recruiters, AI takes over the routine work, so what's left needs more judgment than before. For example, if AI flags the easy scans, radiologists spend more time on complex cases that need real clinical interpretation. That makes their jobs harder to replace, not easier. The same goes for recruiters: if AI screens candidates, the tough decisions are now the main part of the job.

"Democratised" jobs are the opposite. AI makes it easier for more people to do these jobs well, so the extra pay for expertise goes down. If anyone with the right software can do a decent job in IT service management or medical transcription, salaries for those roles usually fall. That's what the data shows.

The 62 percent wage premium for AI skills is mostly found in professionalised jobs. People in democratised roles aren't seeing the same pay increases, and PwC's data doesn't show that changing soon. Salary growth between the two groups is already splitting by 42 percent.

When hiring, the key question before writing a job description is whether the role will become more or less valuable as AI improves. This question affects pay, career growth, and how hard it will be to keep people in the role.

Advice: Figure out which track the job is on before you write the description. Roles that are becoming more common need different pay and retention strategies than those that are becoming more specialized. Pay matters, but so do career paths and what you tell candidates about long-term growth.

2. Cutting Jobs Doesn't Guarantee Performing Better.

Recap: Tech layoffs in 2026 reached 183,966 workers as of June 16, averaging 1,115 per day, nearly double last year's pace. Fifty-five percent of announcements explicitly cite AI as a driver. The same AI-driven split shows up in how companies are cutting. A May 2026 Gartner study of 350 firms found that companies cutting the most aggressively showed no improvement in financial returns. Two additional independent research groups, MIT and METR, arrived at similar findings. Many companies are making permanent workforce decisions based on AI performance they have not yet measured or, in some cases, have not yet deployed.

The numbers are striking: 183,966 workers have been laid off this year as of June 16, averaging 1,115 per day. That's almost twice last year's rate. In 55 percent of cases, companies say AI is the reason.

But here's the catch: a Gartner study from May looked at 350 companies and found that those making the biggest cuts didn't see better financial results. MIT and METR found the same thing. The idea that cutting lots of jobs boosts productivity sounds good, but the evidence isn't there yet.

There's another problem. Some companies are making big, permanent changes based on AI tools they haven't even rolled out or measured yet. If the expected results don't show up, these companies will have less flexibility to adjust. Sam Altman called this "AI washing" earlier this year, meaning companies blame AI for cuts when the real reasons are things like overhiring or profit pressure. The Gartner data backs this up.

For founders hiring now: most people affected by these layoffs aren't low performers. Earlier layoffs already handled that. Now, experienced product and engineering talent is on the market because of restructuring driven as much by shareholder signals as by business needs. That is the candidate pool to keep in mind when you're reviewing candidates.

Advice: It's worth talking directly to people coming out of this round of tech layoffs. Gartner's research shows many weren't let go because they were replaceable, but because their companies gambled on AI results that haven't materialized. That's very different from someone who was let go for performance reasons.

3. Colombia Votes Sunday.

Recap: Colombia's presidential runoff takes place June 21 between far-right candidate Abelardo de la Espriella, who led the first round with 43.7 percent, and leftist Iván Cepeda, aligned with outgoing President Gustavo Petro, who received 40.9 percent. De la Espriella has drawn comparisons to Argentina's Javier Milei and has campaigned to reverse Petro's economic policies, lower corporate taxes, and reopen Colombia to foreign direct investment. The three-point first-round lead is significant but not decisive in a two-candidate election. That is why it's worth looking back at what we wrote on May 27. For three years, investor caution under Petro kept foreign direct investment 9 percent below 2022 levels, which held down both salaries and competition for top Colombian talent. We said the election would either keep things the same or change them. Shift it.

The first-round results suggest change is coming. De la Espriella got 43.7 percent and Cepeda got 40.9 percent. That's a real gap, but 43.7 percent in a crowded first round doesn't guarantee a win in the runoff. Recent polls show the race is tightening, so Sunday's result is still up in the air.

If you're building teams in Bogotá or Medellín, timing matters. Multinational companies waiting for clear rules will start making moves before the new government takes office. Salary benchmarks and job offers change based on what people expect, not just what's happening now.

The window we talked about on May 27 is still open. But depending on Sunday's result, it could close sooner than the data will show.

Advice: If you've been waiting to start hiring in Colombia, Sunday is your last clear signal before things change. The job candidates you can hire at today's pay rates may not be available at those rates six months from now if the investment climate shifts.

That is it for this week.

The US won 4-1 and Colombia votes in four days. PwC's analysis of one billion job postings shows salary growth diverging at 42 percent between the two tracks AI is creating. The tech layoff data show that the efficiency case for mass cuts is not holding up under scrutiny. The Colombian hiring window we described on May 27 is about to get narrower.

That's what this newsletter is here for. It's also what we help clients with every week at lupahire.com.

Until next time,

Joseph Burns
CEO & Founder, Lupa

Hey there,

The World Cup kicks off tomorrow in Mexico City. There will be 104 matches, 48 teams, and 39 days of games. New research this week says US employers could lose between $17 billion and $30 billion in productivity by the time the final match ends at MetLife Stadium on July 19.

That number is eye-catching, but what matters more is what's driving it, especially for anyone building a team right now.

During the same week as the tournament kickoff, the April JOLTS report showed job openings rising to 7.62 million, the highest in two years. Professional and business services made up 90 percent of that increase. Meanwhile, tech layoffs in 2026 surpassed 142,000, with most companies citing AI as the main reason.

Three data points. Most companies are paying attention to the wrong one.

Let's get into it.

1. The World Cup Opens Tomorrow. For Companies Hiring in Latin America, the Next 39 Days Have a Specific Logic.

Recap: The 2026 FIFA World Cup begins June 11 with Mexico vs. South Africa in Mexico City and runs through July 19, with 104 matches across 16 cities in the United States, Mexico, and Canada. A UKG survey of 8,000 employees across eight countries found that 37 percent of workers plan to adjust their schedules during the 39-day tournament, with 27 percent planning to arrive late, leave early, or take full days off. Separate analysis puts the productivity cost to US employers at up to $30 billion. 19% of employees said they would consider looking for a new job if their employer's scheduling policies interfered with their ability to watch matches.

The productivity number is real, but it's a rough estimate. It just adds up the hours lost from the 37 percent of workers who plan to change their schedules and turns that into dollars. What it misses is where those changes actually happen, which is the more useful question for anyone managing teams in Colombia, Brazil, or Argentina.

Colombia opens on June 17 against Uzbekistan. Argentina starts on June 16, and Brazil on June 13. All these matches take place during the workday in the Eastern and Central time zones. Senior engineers and operations leaders on those teams will be less available on those days. Decisions will take longer, and scheduling interviews around elimination rounds will be tougher. This isn't a criticism of the people, it's just a scheduling reality that most US hiring managers haven't planned for in Q3.

The bigger issue is timing. In Colombia, hiring cycles in professional services already take six to eight weeks. With a major tournament running through July 19, any role that hasn't started its search this week probably won't close until August at the earliest. That means onboarding and ramp-up will likely fall into Q4.

Advice:

If you plan to hire in Latin America this quarter and haven't started yet, you're already behind the tournament schedule. Companies that start now will finish before the final. Those that wait will feel it in their Q3 timelines.

2. Job Openings Hit a 2-Year High. Nobody Is Moving.

Recap: The Bureau of Labor Statistics April 2026 JOLTS report, released June 2, showed job openings rising to 7.62 million, up 731,000 from March and the highest level since May 2024. Professional and business services accounted for more than 90 percent of the increase, adding 668,000 openings. At the same time, the number of new hires fell sharply, layoffs fell, and voluntary quits dropped to their lowest point in almost six years.

7.62 million openings might look like a hot market, but the rest of the report tells a different story.

Job openings show what employers want, not what's actually happening. The April data shows companies posting more jobs, but not hiring more people. Quits are at a six-year low, so workers are staying put even when they have options. Hires and layoffs both dropped. The market is stuck: employers want to hire, employees aren't moving, and actual placements are going down.

This is the same pattern that has made professional hiring tough for the past year and a half. Last week's issue talked about the 31-month white-collar slowdown. This week's data shows the other side. There are plenty of openings, but the best candidates have stayed put through 31 months of cuts because they're good enough to keep their jobs. They won't move just because a new job is posted. Companies waiting for applications are only reaching the small group of people actively looking, not the ones with the skills you really need.

Advice:

7.62 million openings doesn't mean there's a big talent pool. It just shows what employers want. The candidates you really want aren't looking at those postings. They kept their jobs during the downturn and need a clear, direct reason to leave. You have to find them and make your case directly.

3. Tech Layoffs Hit 142,000 in 2026. The Companies Doing the Cutting Are Spending $700 Billion on Infrastructure Instead.

Recap: Tech layoffs surpassed 142,000 in 2026 as of late May and are on pace to exceed 370,000 by year-end, according to industry trackers, making 2026 the largest layoff year in the sector's recorded history. Fifty-five percent of layoff announcements explicitly cited AI or automation as a primary driver. Four hyperscalers, Amazon, Microsoft, Alphabet, and Meta, committed a combined $700 billion in capital expenditure for 2026, nearly double their 2025 spending. GitLab confirmed in June that it cut 14 percent of its workforce, roughly 350 employees, while exiting 22 countries and flattening management by up to three layers as part of what it called a restructuring for the agentic era.

The headline is 142,000 layoffs, but the more important number is $700 billion.

That $700 billion is what four companies are investing in AI infrastructure while running their largest layoff cycles ever. The money isn't going to headcount. It is going to compute. Each company is shifting what it was paying in salaries to what it now pays for servers, chips, and energy contracts.

The people being let go aren't low performers. Companies restructuring for AI have already cut the easier roles. Now, experienced product, engineering, and operations talent who held on through earlier rounds are caught in a restructuring unrelated to individual performance. GitLab's decision to leave 22 countries wasn't about poor performance; it was because centralized AI infrastructure made having teams everywhere less necessary.

That talent is available right now. Most aren't flooding job boards, because those who survived earlier cuts still believe they can find something without making it public. They're right, but only if someone reaches out to them first. The window is short. Six to twelve weeks after a layoff, the best candidates have already accepted new roles.

Advice:

The best candidates from this round of tech layoffs aren't on job boards. They're in the networks of people you already trust. Now is the time to ask your contacts who they know and reach out directly before those candidates take new roles just because they're available.

That is it for this week.

The World Cup starts tomorrow, and the next 39 days have a unique schedule for anyone building teams in Latin America. Most companies haven't planned for it. The April JOLTS data shows 7.62 million openings, but the quit rate is at a six-year low: employers want to hire, people aren't moving, and the best candidates aren't applying. Meanwhile, 142,000 tech workers have been cut in 2026, and the most experienced are available for only a short time before they quietly take new jobs.

These patterns do not wait for companies to notice them.

That's what this newsletter is for. It's also what we help clients with every week at Lupa.

Until next time,

Joseph Burns
CEO & Founder, Lupa

Hey there,

On the surface, the labor market looks healthy. Unemployment is steady at about 4 percent, and job openings are at their highest in nearly two years. But entry-level opportunities are disappearing.

Recent graduates are facing the toughest entry-level job market in years, and more companies are pausing junior hiring altogether. It seems clear that companies are choosing not to invest in early-career talent.

These same pressures are leading US companies to build teams in Latin America, and, as this week's data show, this trend is accelerating. Colombia's recent election also gave investors a reason to be optimistic.

Let's get into it.

🌐 News Shortlist

1. AI Broke the Entry-Level Job. The Damage Is Concentrated at the Bottom.

Recap: New data shows this is the hardest entry-level job market for graduates since the pandemic. Unemployment for recent college grads aged 22 to 27 is now 5.7 percent, higher than the 4.2 percent rate for all workers, and about 43 percent of new grads are underemployed. The National Association of Colleges and Employers predicts 2026 will be the worst graduate job market since 2020. About 21 percent of companies have already stopped hiring for entry-level roles because of AI, and 47 percent expect to cut these jobs entirely by 2027. Still, some white-collar fields expected to shrink due to AI have actually grown since 2022: there are 7 percent more software developers, 10 percent more radiologists, and 21 percent more paralegals.

Looking closer, AI is not wiping out all white-collar jobs. Instead, it is taking away the junior tasks that used to help young people get their start.

This difference is important because it explains a trend that is often misunderstood. Since 2022, white-collar jobs overall have increased, especially at the senior and mid-levels. The jobs disappearing are those that rely on tasks AI can now do quickly, such as first-draft research, basic coding, document review, and entry-level analysis. Employers are now asking for more experience in roles that used to help people gain it. So, the market wants candidates who already have the skills that entry-level jobs once provided.

For founders, this creates a trap that looks like a savings. Cutting junior hires lowers payroll today and removes the pipeline that produces your senior people tomorrow. The companies freezing entry-level roles are just deferring a cost-saving problem.

Advice: Do not cut your junior hiring just to save money in the short term. Figure out which early-career roles help your team build the skills you will need in a few years, and make sure to keep those positions. Companies that continue to develop talent now will have an advantage as others compete for experienced staff.

2. US Companies Are Accelerating Into Latin America. The Cheap Window Is Closing.

Recap: Over half of US companies now plan to expand their nearshoring in Latin America through 2026, building teams in Mexico, Colombia, Costa Rica, and Brazil. Investment in Mexico alone jumped over 70 percent year-over-year to about 28 billion dollars, and the Inter-American Development Bank reports that foreign investment is rising again in Mexico, Brazil, and Colombia. As the market grows, competition for talent is heating up. Developer salaries in the region are increasing by 8 to 12 percent a year in most places, and by 2026, rates are expected to be 10 to 15 percent higher than in 2025. Still, hiring in Latin America is much cheaper than in the US.

This story is the direct answer to the first one. When the US entry-level ladder breaks, experienced talent gets scarcer and more expensive at home. Latin America is where companies are rebuilding that capacity, and the numbers show they are moving fast.

There are good reasons for this shift. Latin America has strong engineering, product, and operations talent, similar time zones, and costs that are still much lower than in the US. The main thing changing now is the price. Salaries are rising by 10 to 15 percent a year because demand is outpacing supply. Companies that started building teams here two years ago secured talent before salaries jumped. Those starting now will pay more than before but less than in the future.

A common mistake is seeing Latin America only as a way to save money, not as a place to build strong teams. Companies that focus just on the lowest costs often hire poorly, lose employees within a year, and then think the region is not a good fit. The region works well, but bargain hunting does not.

Advice: If you are considering Latin America, act this year and use current salary data, not last year's. Build your team with a long-term mindset, offering real opportunities and reasons to stay, because the talent you want has choices and is aware of it.

3. Colombia's Vote Thrilled Investors. The People You Hire Will Outlast the Result.

Recap: In Colombia's first-round presidential vote on May 31, right-wing outsider Abelardo de la Espriella led with 43.7 percent, ahead of leftist senator Iván Cepeda with 40.9 percent. Since no candidate got over 50 percent, both will go to a runoff on June 21. This result surprised many, as polls had favored Cepeda. Markets reacted right away: Colombian stocks had their biggest jump in over six years, bonds rallied, and the peso strengthened as investors expect a more business-friendly approach from De la Espriella, who supports tax cuts, deregulation, and private investment. Still, analysts warn that volatility will likely continue until the runoff.

I am not Colombian and do not vote there, so these are not my personal views. Many people have asked how the election result will affect team building in the region, so here is what analysts are saying.

De la Espriella's plan is what markets prefer: tax cuts, deregulation, a focus on energy, and a smaller government, plus a respected former finance minister as his running mate. Cepeda's platform would continue the current government's policies, with greater redistribution, greater state involvement, and a continued halt to new oil projects. In the short term, investors are more positive about De la Espriella and more cautious about Cepeda, which aligns with the market's initial reaction. However, there are some concerns. De la Espriella's proposals are seen as bold but lacking detail, and his confrontational style brings its own uncertainty. Colombia's fiscal deficit is expected to widen regardless of who wins. Both candidates want lower interest rates, and investors are watching that closely.

Here is what matters for hiring: politics creates the environment, but talent determines your success. I have seen companies do great work in Colombia under both left- and right-leaning governments. The engineers, product leaders, and operators we work with in Bogotá and Medellín will keep building and adapting, no matter who wins on June 21.

Advice: Do not base your hiring in Colombia on the election timeline. Strong teams have succeeded there under every recent government. Start building relationships now, and let your team, not the news, shape your results.

That is it for this week.

Entry-level jobs are disappearing in the US, and companies that cut them are saving money now but risking a shortage of senior talent later. This shortage is one reason US companies are moving into Latin America, where there is plenty of talent and real cost advantages, although the best deals are fading quickly. Colombia's recent election gave investors something to celebrate, but the real key to success there remains the same: your team.

The trends that shift first are the ones to watch right now. That is the purpose of this newsletter, and it is what we help clients understand every week at lupahire.com.

Until next time,

Joseph Burns
CEO & Founder, Lupa

Hey there,

The most useful piece of writing I read on AI and work this week was 42,000 words long. It came from the Vatican.

Pope Leo XIV published his first encyclical on May 25 and presented it alongside Anthropic's co-founder. The document is about AI, labor, and human dignity, and it draws a direct line from the Industrial Revolution to now.

The data this week makes that line concrete.

Let's get into it.

🌐 News Shortlist

1. The Pope Published a 42,000-Word Document on AI and Labor

Recap: Pope Leo XIV published "Magnifica Humanitas" on May 25, his first major teaching document, a 42,300-word encyclical on AI, labor, and human dignity. He signed it on May 15, the 135th anniversary of Rerum Novarum, Pope Leo XIII's 1891 encyclical on labor rights during the Industrial Revolution. The document addresses job insecurity from AI, the concentration of technological power, and the risk of treating workers as something to be replaced or optimized out of existence. The Pope warned that treating human beings as "something to be perfected or surpassed" makes it easier to accept that some lives are less useful or less worthy. The encyclical was presented at a Vatican press conference alongside Anthropic co-founder Chris Olah.

The parallel to 1891 is the detail worth taking seriously.

Rerum Novarum was written during the Industrial Revolution, when factories were absorbing rural workers and the question of what ownership and labor owed each other had no formal answer. Rather than stopping industrialization, Leo XIII's encyclical set out principles for valuing human work within it: the dignity of labor, fair wages, decent working conditions, and the right to organize. Companies that ignored that framework faced a different kind of risk than those that engaged with it.

Leo XIV is making the same move 135 years later. His document asks companies to account honestly for what they are doing to the people whose jobs are changing or disappearing in the process. The line he draws is pointed: if you treat workers as variables to optimize away, you have already accepted a logic that assigns different values to different human lives. That is a harder position to defend than most AI restructuring announcements have acknowledged.

The week before, Standard Chartered's CEO described cutting 7,800 support jobs as "replacing lower-value human capital with investment capital." That is exactly the kind of thinking the encyclical challenges. Even if you do not share the Pope's religious views, his point stands. There is a real gap between how companies talk about AI changes and how workers experience them. A 42,000-word document that tackles this directly deserves more attention than most technology policy writing produced this year.

Advice:

If you cannot explain your AI and staffing decisions in terms of how they affect people, not just efficiency, you do not have a real strategy. You have a narrative. The encyclical calls for honesty about what is really happening. The goal is not to slow down. It is to be honest about the impact.

2. The Unemployment Rate Is Fine. White-Collar Payrolls Have Contracted for 31 Straight Months.

Recap: White-collar jobs have shrunk every month since October 2023, the longest sustained contraction in professional employment outside of a formal recession on record. The headline unemployment rate has stayed near 4.3 percent, masking what has happened in professional services, information, and financial activities. Job openings in these areas dropped below 1 million for the first time since April 2020, according to January 2026 JOLTS data. The average professional job search now takes six months. About 20 percent of job seekers have been looking for ten months or more, and some have sent over 1,700 applications without hearing back.

The headline unemployment rate shows how many people want a job but cannot find one. It does not reveal which sectors are hiring or how long it takes to land one. The 4.3 percent rate has been steady for months. So has the 31-month contraction in white-collar jobs. Both are true at once.

White-collar job losses show that recovery has been uneven. Health care and transportation are adding jobs, but core professional fields like software and finance have seen nearly three years of decline while overall numbers look favorable.

The most valuable white-collar professionals held onto their roles throughout the long downturn because they proved indispensable to their employers. You will not find them on job boards or among the flood of applicants.

Advice:

The 4.3 percent unemployment rate tells you very little about the people you actually want to hire. The best candidates in professional services and tech have made it through a long contraction and are not waiting for your job posting. You need to reach out to them directly.

3. Colombia Votes Sunday. Here Is What That Means for Building a Team There.

Recap: Colombia holds its first-round presidential election on May 31, four days from now. Three candidates are running: Ivan Cepeda of President Gustavo Petro's Historic Pact, Abelardo de la Espriella, a far-right independent drawing comparisons to Argentina's Javier Milei, and Paloma Valencia, a center-right senator from the Democratic Center. A candidate needs more than 50 percent to avoid a June 21 runoff. Foreign direct investment in Colombia is 9 percent below its 2022 level. Investors have been in a sustained wait-and-see posture ahead of the vote, reflecting three years of regulatory uncertainty, high corporate taxes, and a strained relationship with US investors under President Petro.

Beneath the political headlines, this is a story about investment. And that investment story is also a hiring story.

When investor confidence in a country drops, competition for senior talent drops with it. The companies most likely to hire experienced engineers, product managers, and operations leaders in Bogota and Medellin are multinational employers with confidence in the regulatory environment. Three years of investor caution under Petro has kept that competition lower than it would otherwise be. US companies hiring in Colombia right now face fewer competing offers than they would in a fully open investment climate.

That window is directly tied to Sunday's result. A Cepeda win keeps current policies in place, foreign investment likely stays cautious, and talent competition remains low. A Valencia or De la Espriella win signals a more business-friendly posture, accelerating investment and bringing more competition for the senior professionals you are trying to hire.

Most US companies building teams in Colombia have not thought about this election at all. That is a competitive advantage, but it has an expiration date.

Advice:

If you have been planning to build a team in Colombia and have been moving slowly, now is the time to pay attention. If investment picks up after the election, salaries and competition for talent will rise before it shows up in any hiring report. Moving before that happens is the advantage that closes first.

That is it for this week.

The Pope released the most direct moral accounting of AI's impact on workers that any major institution has produced this year. The unemployment rate has been accurate, and so has the 31-month contraction in white-collar jobs underneath it. And Colombia votes in four days, which will shift hiring conditions in Bogota and Medellin before most companies notice.

The trends that change first are the ones worth watching now.

That is what this newsletter is for. And it is what we help clients figure out every week at lupahire.com.

Until next time,
Joseph Burns
CEO & Founder, Lupa

Hey there,

This week, The Globe and Mail noted AI-generated resumes are now so common that jobs can receive over a thousand applications, making it hard for hiring managers to tell authentic candidates from AI-edited ones. At the same time, Meta transferred 15,000 employees in a week to build its AI infrastructure. The tools meant to help are now straining the process, but it's important to look at the bigger picture.

MercadoLibre also reported 49 percent revenue growth, its fastest in four years. This success came from engineers and product managers who any company can hire in Latin America.

The pattern looks familiar. The companies that are hiring well right now are not necessarily the ones using more AI in their processes; instead, they are the ones using better judgment about where to look and what to ask. That is getting harder as the noise grows, and more valuable as it does.

Let's get into it.

🌐 News Shortlist

1. Meta Cut 8,000 and Moved 7,000 More. Now You Can See the Org Chart.

Recap: Meta executed 8,000 layoffs on May 20, with notices delivered at 4 a.m. local time. Two days earlier, the company had told 7,000 employees that they would be moved into four newly created AI-focused organizations covering AI products, applications, and infrastructure. More than 15,000 Meta employees were repositioned in a single week. The company's 2026 capital expenditure plan has been raised to as much as $145 billion, with almost all of it directed to AI infrastructure. Cuts were concentrated in business operations and middle management. The roles being created are specifically focused on building and deploying AI tools.

This week, unlike past layoff headlines, gives you a clear view of what the reorganization actually leads to.

Most focused on the 8,000 layoffs, but the 7,000 internal moves signal how Meta is changing. Those leaving had jobs replaced by AI, while those staying are building AI systems. Meta shifted over 15,000 employees in a week — a complete redesign.

If you're a founder, the real question isn't how many people Meta let go, but what the company looks like now. Meta has 7,000 people in four new AI divisions, set up to work differently. The $145 billion for infrastructure makes this possible, and the 8,000 cuts help pay for it.

Of course, this doesn't guarantee success. Big reorganizations are risky, and AI tools might not take over tasks as smoothly as planned. Meta can afford to take these risks and test the model, but most companies can't.

Advice:

Before you reorganize your team for AI, be clear about which tasks these tools can handle well right now, not what they might do in the future. Companies that get this wrong often reorganize based on hope rather than reality.

2. AI-Generated Resumes Are Breaking the Hiring Process

Recap: A survey of 1,500 hiring managers found that 61 percent say AI-generated resumes are slowing the hiring process, and 89 percent say the volume has increased their workload. Separately, 70 percent of job seekers now report using AI in some form during their search, and 31 percent use it to generate or rewrite their resumes directly. Applications per open role have more than doubled since 2022. Competitive tech and remote roles now routinely attract more than 1,000 applications. Despite that volume, only 4 to 6 candidates typically reach the interview stage.

We can certainly say now that, besides changing who gets laid off, AI has also changed who applies for jobs and what their applications look like.

Candidates use AI to apply faster and to more jobs, swelling application numbers. Recruiters respond by tightening filters or slowing down, causing candidates to apply even more, fueling a cycle where hiring managers still struggle to find quality candidates.

If you're hiring now, this is a real problem. An average candidate with a polished AI-generated resume can look just like a top candidate at first glance. Most screening systems weren't built for a world where almost a third of applicants use the same AI tools.

The underlying problem is the model. Inbound hiring was always a blunt instrument: post a role, wait for applications, filter on what comes in. It worked when the signal-to-noise ratio was manageable. It does not work when 70 percent of applicants use AI to apply, and 31 percent let AI write their applications. No filter corrects for that.

Since the noise is structural, the obvious answer is what successful companies have always searched for: proactive recruitment. Searching and identifying candidates who are not looking, reaching out directly, and building interview processes specific enough to reveal actual capability rather than resume quality. A tailored technical assessment, a conversation designed around the actual problems a role involves, a referral from someone who has worked alongside the candidate — these are hard to game at scale. An inbound application is increasingly not.

Advice:

If you start your hiring process by looking at resumes, you're starting with the least reliable signal right now. Referrals, work samples, and direct outreach will help you find better candidates faster.

3. MercadoLibre's Best Quarter in Four Years Was Built With Latin American Talent.

Recap: MercadoLibre reported Q1 2026 results on May 7. Revenue reached $8.8 billion, up 49 percent year over year, the company's fastest growth rate since Q2 2022. Total payment volume crossed $87 billion, up 50 percent. Mercado Pago's monthly active users increased 29 percent. The credit portfolio nearly doubled to $14.6 billion. In Brazil, sales of items jumped 56 percent. The company now serves customers across 18 countries in Latin America.

MercadoLibre was founded by people who studied at, among others, the Universidad de Buenos Aires, the Pontificia Universidad Católica de Chile, and the Universidad de los Andes. They built a $14.6 billion credit portfolio and a payment system processing $87 billion in quarterly volume. They did it in the same time zones as New York and Chicago, in markets where building is hard, and infrastructure is less forgiving than in Silicon Valley.

The engineers and product managers behind these results haven't aged out of the market. The universities that trained them are still producing new graduates. Professional communities in Bogotá, Buenos Aires, São Paulo, and Santiago are still turning out people with the same skills, and most are available to hire remotely for roles outside MercadoLibre.

Most US companies approach Latin American hiring the same way they approach every other market: post a role, wait for applications, filter on familiar signals. That approach misses most of the best people, who are employed, not actively looking, and not reaching out to companies they have never heard of.

Advice:

The engineers who drove a 50 percent surge in payment volume are available to hire. Most US companies do not know where to look or how to reach them. That gap is the competitive advantage, and it does not stay open forever.

That is it for this week.

Meta demonstrated how to redesign an organization around AI with clear action. The Globe and Mail showed that relying on resumes is an increasingly unreliable way to find strong candidates. MercadoLibre revealed that high-performing technical talent is available in Latin America, waiting for companies that know how to access it. The main takeaway: rethinking where and how you hire is now a competitive advantage.

The clear pattern: companies succeed when they spot these trends and adapt their hiring and organizational strategies accordingly.

That is what this newsletter is for. And it is what we help clients work through every week at lupahire.com.

Until next time,
Joseph Burns
CEO & Founder, Lupa

Hey there,

I have been getting versions of the same question from clients lately: "Okay, but are the AI layoffs actually real, or is AI the excuse?" They have read the skeptical takes, including in this newsletter. And they are starting to wonder if they are being too dismissive.

The honest answer this week is: one of them probably is.

Cloudflare cut 20 percent of its workforce on May 8, its first layoffs in 16 years, and the CEO said the quiet part out loud. Without quoting "optimizing for efficiency," or "investing in the future." Instead, he said AI usage at the company grew 600 percent in three months, and those specific jobs no longer exist in the same form. The company posted record revenue in the same announcement.

That is a different story from what Big Tech has been telling us. It deserves a different read.

Let’s get into it.

🌐 News Shortlist

1. April Added 115,000 Jobs. February Was Actually Worse Than We Thought.

Recap: The Bureau of Labor Statistics released the April Employment Situation on May 8. Employers added 115,000 jobs in April, more than double the 55,000 to 65,000 economists had forecast. The unemployment rate held at 4.3 percent. Health care led all sectors, followed by transportation and warehousing, and retail trade. Average hourly earnings rose 3.6 percent year over year. Federal government employment continued to decline. February's numbers were revised down to -156,000, from the initially reported -133,000. March was revised up to 185,000.

The headline number is good. Two consecutive months of upside surprises say the labor market has not broken the way many predicted. But the revision to February is the part worth sitting with. The month we thought was bad turned out to be significantly worse. This is how labor market data works: the headline runs everywhere on release day, and the revisions quietly change the underlying picture a few weeks later.

What the April report actually shows is a market that is adding jobs, but not evenly and not across the sectors most relevant to founders. Health care and transportation are structural. Retail reflects consumer spending that may or may not hold. The federal workforce is still contracting, and that talent is still entering the private market. If you are hiring in operations, program management, or policy, that pool is real and still underutilized.

The sectors that are not hiring, professional services, information, and financial activities, are where most people reading this newsletter do their business. For software, marketing, analytics, and operations roles, the April report tells you less than it seems. The market there is still tight in the ways that matter: the candidates you want are employed, cautious about moving, and getting more selective about what it takes to leave.

Advice:

The 115,000 number is real but narrow. If you are hiring in logistics, health care, or operations, conditions have improved. If you are hiring in professional services or tech, the headline does not tell you much about the talent you actually need.

2. Cloudflare Said the Quiet Part Out Loud

Recap: On May 8, Cloudflare announced it was laying off more than 1,100 employees, roughly 20 percent of its workforce, in the first mass layoff in the company's 16-year history. CEO Matthew Prince and President Michelle Zatlyn published a blog post explaining that AI usage inside Cloudflare had grown by over 600 percent in three months, with AI agents now handling tasks across HR, finance, marketing, and engineering. The CEOs stated explicitly that the cuts were not a cost-cutting exercise. Cloudflare reported record quarterly revenue the same week. Other companies announcing cuts the same week included Upwork, which cut 25 percent of its workforce; Coinbase, which cut 14 percent of its workforce; and PayPal, which announced plans to reduce headcount by 20 percent over the next two to three years.

Every other major AI-framed layoff we have covered this year has had a financial story underneath it. Oracle needed to fund $156 billion in infrastructure. Meta and Microsoft were converting payroll into AI capex. Amazon cut while AWS grew 24 percent. The pattern was clear: these were companies repricing labor to fund a capital spending race, and AI was the narrative that made it acceptable.

Cloudflare is different. It is a company with roughly 5,000 employees, not 100,000. It is posting record revenue, so financial pressure is not the explanation. And the CEO's statement is more specific than anything we have seen from a tech executive this year. Not "AI will reshape work over the next decade" but "AI usage inside our company grew 600 percent in three months, and these tasks no longer require the same headcount." That is a different claim. It might still be overstated. Executives describe their decisions in the best available terms. But the specificity is harder to dismiss.

What changes this week is the scale of the company making the claim. Big Tech making AI automation arguments is easy to discount because the financial incentives to make that argument are so large. Cloudflare, with 5,000 people, is closer to the companies that most founders are building or working at. When a company of that size cuts 20 percent during record revenue and attributes it specifically to internal adoption, it is worth treating as a real signal rather than a PR move.

The same week also brought cuts from Upwork, Coinbase, and PayPal, none of them Big Tech, all citing AI. The pattern is moving down market.

Advice:

Cloudflare is the first company at a relatable scale to have tied headcount cuts to documented internal AI adoption rather than to capital reallocation. If you have not audited which workflows AI tools now handle inside your own company, this week is a reason to do it before someone else's board asks you the same question.

3. Nubank Reports Today. The Number Matters More Than the Stock Price.

Recap: Nu Holdings, the parent company of Nubank, reports its first-quarter 2026 earnings today, May 14. Analysts project revenue of approximately $5 billion, representing nearly 53 percent year-over-year growth. Earnings per share are expected to be $0.20, a 67 percent year-over-year increase. The company ended 2025 with 131 million customers, making it the largest private financial institution by customer count in Brazil and a leading fintech across Mexico and Colombia. Nubank launched in 2013 with a credit card and no branches; it now serves roughly 15 percent of Brazil’s adult population.

The Nubank earnings report matters for a reason that has nothing to do with whether you own the stock.

When a company grows revenue by 53 percent annually and has 131 million customers, it is consuming a large amount of senior talent. Engineers, data scientists, product managers, risk analysts, and operations leads are all competing to work there. What Nubank pays, what it offers in terms of scope and growth, and what it expects from its people becomes the reference point for the senior tech labor market in Brazil. That is why these numbers matter to founders hiring remotely in São Paulo or Belo Horizonte, even if their company has nothing to do with fintech.

In the same week that US companies are announcing tens of thousands of job cuts, one of the world’s largest fintech companies is reporting 53 percent revenue growth, driven almost entirely by engineers, product managers, and operators trained in Brazil and Mexico. Nubank was not built by importing people from Silicon Valley. It was built by graduates of UNICAMP, UNAM, and Universidad de los Andes who chose to build something serious in their own region.

For founders who treat Latin America as a secondary market for affordable labor, this is worth revisiting. The talent pool that built Nubank is the same pool you are competing for.

Advice:

If Nubank's numbers come in strong today, expect salary expectations among senior technical professionals in Brazil to keep rising through 2026. Set your compensation benchmarks against what leading regional companies are actually paying now, not against what US companies paid two or three years ago.

That is it for this week.

Three stories that reward a closer read than the headline suggests. The jobs report looked good, but the revision and the sector breakdown tell a more complicated story. Cloudflare's cuts are the first of this cycle that appear to be about actual AI adoption rather than financial engineering, and that company is closer to your size than Meta is. Nubank's earnings are a proxy for how competitive the Latin American senior talent market has become, whether or not you work in fintech.

The talent decisions that hold up over the next two years will be made by people who read the revision, not just the headline.

That is what this newsletter is for. And it is what we help clients work through every week at lupahire.com.

Until next time,
Joseph Burns
CEO & Founder, Lupa

Hey there,

I keep seeing stories of people who laid off their devs and now spend more money on AI tokens than they did with a senior team.

AI itself is not what drives up costs. The real issue is a lack of structure. If you let go of your team without a clear plan for how tools will replace their judgment, knowledge, and work, you end up having to rebuild what you lost. It just costs more, and fewer people remember why things were done a certain way.

The companies that succeed are the ones that understand what AI can really do, where people are still needed, and how to train their teams to work with both. This takes more planning than just sending out a memo about layoffs, but you do not need to be as big as Meta or Amazon to make it work.

At Lupa, we are having this conversation with clients more than any other right now. The real question is how to build a team that knows how to use these tools so well that everyone becomes more valuable than before.

Let's get into it.

🌐 News Shortlist

1. The March Hiring Surge Got Buried Under the Noise

Recap: The Bureau of Labor Statistics released its March Job Openings and Labor Turnover Survey on May 5. Job openings stayed at 6.9 million. Hires rose to 5.55 million, up 655,000 from February, the biggest monthly gain in over a year. Professional and business services added 165,000 hires. Transportation and warehousing added 108,000. Accommodation and food services added 124,000. Quits and layoffs stayed about the same.

Even though headlines focused on AI layoffs and tariff worries, employers actually hired at the highest monthly rate in over a year. That is what the JOLTS numbers show, but it did not get much notice.

The fact that professional and business services led the hiring surge is important. Companies hire in this area when they are planning for the future, not just filling empty spots. Adding 165,000 hires in one month shows these were thoughtful decisions, not just routine changes.

The broader picture is more complicated. Job openings remain at 6.9 million while monthly hires are still well below that figure, meaning companies are looking but taking longer to commit. Q1 GDP came in at 2.0 percent on April 30, below the 2.3 percent consensus, with most of the growth driven by a government shutdown reversal and defense spending rather than organic private-sector momentum. The macro has weaknesses that the headline number does not show.

Still, the JOLTS data speaks for itself. Employers who hired in March did not wait for a better GDP number. They made their call based on what their businesses needed. If you are still waiting for conditions to feel safer before hiring, the data shows that a meaningful number of your competitors have already acted.

Advice:

Professional and business services hiring moved in March, while most people were looking elsewhere. If you have an open role you have been holding until things settle, the market around you is not waiting.

2. Big Tech Is Repricing Labor. It Is Calling That AI.

Recap: Amazon, Meta, and Microsoft cut or restructured tens of thousands of roles in early 2026, all citing AI-driven efficiency as the primary justification. Meta announced 8,000 cuts effective May 20, concentrated in recruiting, HR, and Trust and Safety. Amazon eliminated approximately 16,000 corporate roles in Q1 while reporting AWS growth of 24 percent, its fastest in 13 quarters. Microsoft offered voluntary separation packages to roughly 8,750 US employees. Across recent earnings calls, Meta and Amazon executives referenced efficiency fifteen times combined. Collectively, Google, Amazon, Meta, and Microsoft plan to spend $725 billion on AI capital expenditure in 2026, up 77 percent from last year. A Bloomberg analysis found that roughly half of the roles attributed to AI-driven cuts will be rehired offshore or at lower salary bands.

The AI narrative is carrying a lot of weight in these announcements. It sounds inevitable and strategic, which is easier than saying we overhired during the pandemic, and the balance sheet needs correcting. The Bloomberg finding complicates the story. If half the cut roles are rehired elsewhere at a lower cost, this is as much a labor repricing story as an automation one.

What is actually happening is a specific financial trade-off. Companies spending $725 billion on AI infrastructure need to find that capital somewhere. Payroll is the highest controllable cost and the one that generates the most useful narrative when explaining the decision to shareholders. Efficiency is accurate if you define it as the conversion of people costs into infrastructure costs. It is not the same as proving that software made those jobs unnecessary.

This difference is important for founders. Big Tech can make these moves because of its size. If Meta cuts three thousand engineers, they still have thousands left and a huge research budget. But if a fifty-person company cuts one engineer to pay for a new tool, their whole product plan could change for months. Smaller companies do not have the same flexibility.

The Washington Post's May 1 analysis says the same thing: these layoffs are not just about AI. Some are about stock compensation timing, others about product changes, and some are about headcount that should have been managed sooner. Framing it as an AI story makes it easier for the market to accept, but that does not make it true.

Advice:

When Big Tech talks about AI-driven efficiency, think about what they are not saying in their earnings calls. For your own team, the real question is not how many people to cut, but whether you have the right people.

3. A Twenty-Five-Year Trade Deal Just Changed the Latin America Investment Map

Recap: On May 1, 2026, the EU-Mercosur free trade agreement entered provisional application after 25 years of negotiations. The agreement creates a 700-million-person trading zone and is the European Union's largest trade deal in terms of the scope of tariff reductions. It sharply reduces or eliminates tariffs on cars (previously as high as 35 percent in Mercosur countries), machinery, pharmaceuticals, and agricultural goods in both directions. EU firms will save an estimated $4.68 billion annually in tariff costs. The deal came into effect as US tariffs under the current administration have pushed both regions to diversify trade relationships.

Brazil, Argentina, Uruguay, and Paraguay have just been connected to the EU’s single market at a significantly lower cost. That is not only an agricultural export opportunity. It is an investment story. European companies now have a cleaner, more economically attractive path to establish operations inside Mercosur. They will use it.

The EU-Mercosur deal accelerates a trend already underway: companies are moving money out of US-focused supply chains because tariffs are making them more costly. European manufacturers, logistics firms, and tech companies seeking nearshore options to serve both North America and Europe will increasingly turn to Brazil, Argentina, and the rest of Mercosur. This will bring more foreign investment, more multinational jobs, and more competition for experienced professionals in those markets.

For US founders hiring in Latin America, the window before European companies arrive in force is narrowing. Today, you compete primarily against local employers and other US companies for senior technical and operational talent in Brazil and Argentina. In a few years, you will also be competing against German industrials, French logistics firms, and Spanish financial services companies that now have a trade framework, making their presence in Latin America significantly more economical. That competition is not here at full scale yet, but it is coming.

Advice:

The EU-Mercosur deal is a long-term change, not just a short-term news story. But the demand for experienced talent in Latin America will rise before it shows up in hiring numbers. If you are planning to build a team in Brazil or Argentina, moving now puts you ahead of the curve.

That is it for this week.

Three stories, one pattern: the headlines do not tell the whole story. The JOLTS data show hiring picked up in areas that the layoff narrative had frozen. The Big Tech AI story is really about finances, not just AI. And a trade deal signed twenty-five years ago just changed the economics of investing in Latin America, even if it will take time to show up in hiring numbers.

The best hiring decisions get made with a clear view of what is actually happening, not what is making headlines. That is what this newsletter tries to bring every week. It is also what we help clients do at lupahire.com.

Until next time,
Joseph Burns
CEO & Founder, Lupa

Hey there,

Lately, I’ve been thinking a lot about my time in Latin America and the impact, positive or negative, that we gringos have here. Many foreigners move to cities and skip Spanish lessons. They pay rent in dollars and push up prices for locals. They expect the city to adjust to them. That works for a while, but then it stops. Locals notice when people take from a place without giving back. Sooner or later, there’s a price to pay.

I’ve been here long enough to see both sides. Early on, I had to ask myself: am I just taking advantage of a good market, or am I building something that gives back?

That question is a big part of why I started Lupa. I wanted to connect US companies with Latin American talent in a way that really works. When companies treat hiring here as just a way to save money rather than a real commitment, they usually fail within a year. They blame the region, leave, and leave the people they hired in a tough spot. That hurts everyone: the company, the hires, and the market.

The goal was always simple: help companies succeed in Latin America so they want to return, and place candidates in jobs where they can really grow. When both sides win, everything works.

Let’s get into it.

🌐 News Shortlist

1. The Hiring Market Looks Active. It Isn’t.

Recap: New analysis from Indeed Hiring Lab, published April 13, shows a widening gap between job postings and completed hires in the US. Job postings have climbed steadily, but the actual hiring rate has fallen from 4.5% in 2021 to 2.8% today. Supporting data from BambooHR’s 2026 State of Hiring report shows applicants per posting have nearly doubled, from roughly 46 to 95 over the same period. Worker productivity rose 4.9% last year, one of the sharpest increases on record. At a late-2025 CEO summit, two-thirds of top executives said they plan to keep headcount flat or reduce it, citing trade policy uncertainty, elevated interest rates, and AI.

The US labor market isn’t falling apart. It’s just stuck, and that difference matters to founders trying to hire right now.

Here’s what’s really happening: companies post jobs to keep their pipelines full and look stable, but they aren’t making offers like they used to. Productivity gains that should mean more hiring are instead used as reasons to keep team sizes the same. Teams that once needed eight people now do the same work with six, and leaders are fine with that. So jobs stay open, applications pile up, and nothing gets filled.

This sets a trap for founders. The market looks competitive because there are so many applicants, but the people you really want are mostly still employed and not moving. Quit rates are near record lows. The best candidates aren’t actively looking. They feel secure enough to wait, especially since tariffs and AI make changing jobs riskier than it was two years ago.

This caution makes the freeze worse. Employers aren’t investing in their teams as much as the headlines suggest, and top candidates aren’t taking big risks. The whole market is waiting for things to settle.

Advice:

If you have an important opening, treat it like a real competition, even if you get lots of applicants. High volume doesn’t mean high quality. The people you want need a strong reason to move, and right now, that bar is higher than it’s been in years.

2. Meta and Microsoft Cut 23,000 Jobs and Blamed AI. That Is Not the Whole Story.

Recap: On April 23 and 24, Meta and Microsoft announced cuts affecting a combined 23,000 employees within a 24-hour window. Meta plans to eliminate 8,000 roles starting May 20, concentrated in its Trust and Safety division, where AI moderation tools have reportedly surpassed human reviewer accuracy across most content categories. Microsoft announced buyouts to approximately 8,750 US employees. Both companies explicitly cited AI-driven efficiency gains as the primary driver. The announcements came the same week that Alphabet, Amazon, Meta, and Microsoft confirmed plans to spend nearly $700 billion combined on AI infrastructure in 2026.

Both companies do have real business reasons for these cuts. AI moderation tools really are replacing some content review jobs. That part is true.

What deserves more scrutiny is the framing. A December 2025 survey found that 59% of hiring managers admitted they emphasize AI when announcing layoffs because it plays better with stakeholders than acknowledging overhiring or margin pressure. “AI did it” quickly closes the conversation. It sounds like an inevitable force of nature rather than a deliberate set of choices made over several years. It shifts accountability away from the decisions that led to the overcorrection in the first place.

This matters for founders paying attention to these headlines and trying to decide what to do next. The idea that AI is wiping out whole business functions is only partly true—and it’s also helpful for those who want that story to be believed. The same companies cutting jobs are spending more than ever on AI infrastructure. That’s a choice about where to put their money, not proof that the labor market is falling apart.

A better signal is what these companies are hiring for now: AI integration roles, technical customer success, forward-deployed engineers, and enterprise AI support teams. For every job lost to automation, another is growing to manage, deploy, or sell it. The shift is real and uneven, but it’s not a total elimination.

Advice:

Don’t let Big Tech’s reasons for restructuring shape your hiring plans. Most of what they’re doing is correcting for too much hiring between 2020 and 2022. Founders who build teams that fit their real needs and train people on AI tools now will be in a much better spot than those who use AI uncertainty as an excuse to stop hiring.

3. Argentina Got a $1 Billion Vote of Confidence from the IMF.

Recap: On April 15, the IMF cleared Argentina’s second program review, unlocking a $1 billion disbursement and endorsing a package of structural reforms. The package includes labor market flexibility legislation, the restoration of companies’ ability to repatriate profits, and the formal dismantling of capital controls that had been in place since 2019. The IMF projects Argentina will grow 3.5% in 2026, the strongest growth forecast among the region’s major economies. However, March 2026 came in at 3.4% monthly inflation, bringing Q1 cumulative inflation to 9.4%. The government had budgeted 10.1% for the entire calendar year.

Argentina is really recovering. The big picture—FDI coming back, capital controls gone, and the IMF program on track—shows real structural change. For companies that had given up on the country, there’s now a strong case to take another look, backed by real institutional support.

The inflation numbers add important context. The government planned for 10.1% inflation for all of 2026, but they’ve already hit 9.4% in just three months. This doesn’t mean stabilization is failing; it just means costs are rising faster than expected. Companies hiring in Argentina now need to update their view that it’s a very low-cost market. Senior professionals know their buying power is shrinking and are pricing their work to match. The best candidates have been quoted in USD for a while, and that’s not changing.

This creates a more complex opportunity, not a simple one. Argentina’s talent—across engineering, product, finance, and legal—is still very strong, thanks to top universities and a long history of technical skill. That hasn’t changed. What has changed is pay and how candidates think. The best people aren’t hugely undervalued anymore, and many now look for jobs paid in dollars because they know how inflation affects peso salaries. That’s still a real advantage for US companies hiring there, but it takes better preparation and stronger offers than before.

Advice:

If you plan to hire in Argentina in 2026, use up-to-date numbers. The big picture is getting better, but costs are rising, and top candidates expect more. Companies that come in with real pay benchmarks and a clear commitment to the market will build much better teams than those still using 2023 assumptions.

That’s it for this week.

All three stories come back to the same question: are you really building something, or just looking for an easy advantage?

A frozen labor market rewards companies that put in the effort to find and win over the right candidate, rather than just waiting for many applicants. AI layoff stories deserve a closer look—founders who ask, “Who benefits from this story?” will make better choices than those who just accept it. In Argentina, companies that come prepared and make real offers do better than those who expect the market to adjust to them.

One way to hire in Latin America is to take value and leave. Another way is to build something lasting and return. The second way is harder, but it’s the only one that matters. If you want to talk about what that could look like for your team, reach out to us at lupahire.com.

Until next time,
Joseph Burns
CEO & Founder, Lupa

Hey there,

This week, I was named to Atlas and Hoxo’s Top 100 Recruitment Leaders in North America. Honestly, I had no idea I was on their radar.

A year ago, I made a commitment to post on LinkedIn every day. I stuck with it, knowing that consistency mattered—not just for the algorithm, but because you only get better by putting your work out there. Even after 11 million impressions, I never expected this outcome.

I also learned how important it is to listen. The comments, direct messages, and even the pushback all help. That feedback makes the content better over time.

But honestly, the real reason any of this works is simple. I do not focus on myself. I write about Latin America—its talent, history, culture, and the amazing talent markets that many US founders miss. That focus is why I started Lupa. When content is based on real experiences, people notice.

Grateful to represent Latin America recruiting on the leaderboard. And to everyone who reads, comments, and shares, you are the reason this platform means anything. This is just the beginning.

Let’s get into it.

🌐 News Shortlist

1. The March Jobs Report Just Threw Everyone’s Predictions Out the Window

Recap: The US economy added 178,000 jobs in March 2026, far exceeding the Dow Jones consensus estimate of 59,000. The gain follows February’s revised loss of 133,000 jobs. Unemployment held at 4.3%. Health care, construction, and transportation led the gains. Federal government employment continued to decline, with the government now down roughly 355,000 employees since early 2025. Meanwhile, labor force participation slipped, and long-term unemployment remained elevated.

The headline number looks great—three times the forecast and the biggest monthly gain since Trump’s second term began. But if you are actually hiring, the details in the report matter more and tell a more complex story.

Right now, the labor market is best described as ‘low-hire, low-fire.’ Most people are not being laid off, except in the federal workforce, but they are not changing jobs either. Quit rates are at historic lows because workers feel it is risky to switch. While this looks good for retention, it actually means the pool of candidates for open roles is smaller and less active than it appears.

The sectors adding jobs show where demand is strongest. Health care continues to grow, corresponding with demographic trends. Construction is up because of infrastructure projects across the country. The shrinking federal workforce is sending experienced people into the private sector. If you are hiring for operations, compliance, or program management, this is a talent pool worth considering.

Overall, this is not a boom or a bust. The best candidates have choices and are prudent about making moves. If you want to hire top talent now, you need to act quickly and make stronger offers than you did last year.

Advice:

Do not assume the market is opening up just because of the 178,000 jobs number. The people you want are already employed and comfortable. Your offer has to be strong enough to make them consider leaving their current job.

2. 67% of Workers Say Their Employer Is Failing on AI Training

Recap: New research from SHRM and Gloat found that 67% of US workers disagree that their employer has acted proactively in training them to work alongside AI. Only 17% report that their company is doing anything meaningful to upskill workers in AI-impacted roles. Meanwhile, 51% of workers say they want more AI training. Separately, 80% of HR leaders now use AI for functions like researching employment laws and summarizing reports, but only one in ten has fully automated any HR function.

The reskilling gap is turning into one of the year’s biggest hiring stories. It is not a new problem, but it is getting worse just as AI adoption speeds up.

It is simple: companies are rolling out AI tools faster than they are training employees to use them. This creates a two-speed workforce—a small group of ‘power users’ who learn on their own, and a much larger group that falls behind. As Anthropic’s head of economics said at the Axios AI Summit last month, there is no big difference in unemployment figures between AI-exposed and non-exposed jobs yet. But the skills gap within those jobs is growing quickly.

This has a direct impact on hiring. If you are a founder or operator building an AI-ready team, you have two choices. You can hire people who already have these qualifications and compete with everyone else, or you can hire strong candidates and train them. Right now, 67% of workers say their employer is not investing in training. That is your chance.

The EU AI Act is also worth mentioning here. By August 2, 2026, any company using AI to screen, rank, or evaluate job candidates in the EU must comply with documentation, bias audit, and human review requirements. If you use AI in your hiring process and have any European exposure, that deadline is four months away.

Advice:

Companies that train their teams on AI now will have an advantage in hiring and keeping talent over the next two years. Those who wait will be fighting for the same small group of self-taught people. Build these skills inside your company—it is cheaper, faster, and helps build loyalty.

3. Foxconn Bet $900 Million That Mexico Is the Future of AI Hardware

Recap: Foxconn is building a massive AI server assembly facility in Jalisco, Mexico, with an investment of approximately $900 million. The plant will produce NVIDIA GB200 NVL72 servers, primarily for OpenAI’s $500 billion Stargate project. NVIDIA’s sales in Mexico grew 300% over the past year, and the company has added 45 new corporate partners in the country since late 2024. Foxconn recently invested an additional $136 million to expand capacity and was not slowed by tariff uncertainty.

While the US discusses tariff refunds and return-to-office rules, Foxconn is busy building the world’s largest assembly operation for advanced AI servers—in Mexico.

The Jalisco plant will make NVIDIA’s GB200 servers, which power the Stargate project. Foxconn did not pause when tariffs were announced. The company grew its server business in Mexico during Trump’s first term and is investing even more now. When asked about tariff risks, several manufacturers gave the same answer: they are staying.

It is not exclusively about cost. Geography, mature supply chains, and a growing talent pipeline all play a role. NVIDIA is working with UNAM and Tec de Monterrey. There are now 1,100 AI startups in Latin America using NVIDIA’s platforms. Mexico City has the region’s largest tech talent pool, with 300,000 professionals.

If you think of Latin America only as a nearshoring option, it is time to broaden your view. The region is not simply a place to hire software engineers—it is where some of the world’s most advanced hardware is being built for the biggest AI project ever announced. The talent pool and infrastructure are growing side by side.

Advice:

Latin America’s value is rising fast. If you have considered the region for engineering or operations hires, now is the time to move. The companies investing $900 million are not making guesses—they have already run the numbers.

That’s it for this week.

Three stories, one theme: the labor market is not broken, it is changing. New jobs are appearing, but they need new skills. Companies are putting millions into new regions, but talent pipelines are still developing. The gap between companies that adapt and those that wait is growing every quarter.

The best thing you can do now is simple: hire well, train your team, and build where the fundamentals are going. That is exactly what we help companies do at lupahire.com.

Until next time,
Joseph Burns
CEO & Founder, Lupa

Hey there,

Happy April Fools’ Day. I almost started with a fake headline, like “Lupa raises $500M Series A” or “Every tech company agrees to stop calling layoffs ‘organizational realignments.’” Honestly, though, this week’s real headlines are already unbelievable.

Oracle laid off up to 30,000 people yesterday morning via a 6am email. Meanwhile, LinkedIn data shows AI has created 1.3 million new jobs in two years. And the trade agreement that makes cross-border hiring in North America possible is up for its first formal review this summer.

The real news is more interesting than anything fake. Instead of “AI is destroying everything,” the real story this quarter is that the labor market is changing quickly, and companies that adapt will come out ahead.

At Lupa, we spent the week doing what we always do: pairing the right person to the right role. That stays the same, no matter how surprising the headlines are.

Let’s get into it.

🌐 News Shortlist

1. Oracle Fires 30,000 People With a 6am Email

Recap: On Tuesday, March 31, Oracle began what analysts believe could be the largest layoff in company history. Employees across the US, India, Canada, and Mexico received termination emails from “Oracle Leadership” at approximately 6am local time with no prior warning. Investment bank TD Cowen estimates the cuts will affect 20,000 to 30,000 employees, roughly 18% of Oracle’s global workforce. Access to company systems was cut immediately.

The numbers are big. Oracle announced a $2.1 billion restructuring plan in its SEC filing. TD Cowen says the layoffs will free up $8 to $10 billion in cash flow, which will go straight into AI data center infrastructure. Oracle’s net income jumped 95% last quarter, and it has $523 billion in remaining performance obligations. Financially, this firm stays stable.

However, what people will remember is how it was done. No phone call, no meeting, no talk with a manager. Just a 6am email and a locked laptop. Employees on Reddit said some teams lost at least 30% of their members, and some people found out they were let go only when they lost access before even reading the message.

There are business reasons to restructure when investing. Companies change direction and strategies regularly. That’s normal. However, treating the process like a logistics task—where 30,000 people learn about their jobs from an automated email before sunrise—is not. How you handle a transition shows your remaining employees, your customers, and the market who you really are.

Oracle will likely be fine financially. The $156 billion infrastructure build is a real strategy for real demand. However, the knowledge those 30,000 people had doesn’t move to a data center. And the trust of the employees who are still there just took a hit that no all-hands meeting can repair.

Advice:

If you ever need to make cuts while building a company, plan how you’ll communicate it before you plan how much you’ll save. How you deliver the news matters more than the memo itself.

2. AI Has Quietly Created 1.3 Million New Jobs

Recap: A LinkedIn report published in partnership with the World Economic Forum found that AI has created more than 1.3 million new jobs globally in just 2 years, including roles such as AI Engineer, Forward-Deployed Engineer, and Data Annotator that barely existed before 2023. The WEF projects 170 million new roles by 2030, with a net gain of 78 million jobs after accounting for 92 million displaced. LinkedIn also reported a 34% year-over-year increase in AI/ML job postings, even as overall tech postings declined 8%.

This is the story that deserves more attention. It not only contradicts the layoff headlines but also further adds important context to them.

The labor market isn’t falling apart—it’s shifting. Global hiring is still about 20% below pre-pandemic figures. Nevertheless, LinkedIn’s data shows the slowdown is mostly due to economic insecurity and post-pandemic adjustments, not AI taking over. Except for a few roles, jobs with high and low AI exposure are being filled at about the same rate.

What’s changing is the mix of jobs. AI Engineer is now LinkedIn’s fastest-growing job title, up 143% from last year. The typical hire has 3.7 years of experience, so it’s not just for senior executives. Mid-career professionals with the right skills can get these roles. Plus, 52% of professionals worldwide say they’re actively job hunting in 2026. The talent and demand are both there—the match just looks different than it did two years ago.

The WEF gave a number worth remembering: a net gain of 78 million jobs by 2030. The transition won’t be easy; nevertheless, the opportunity is real for people and companies set to adapt.

Advice:

If you’re hiring, don’t look for the same profiles you hired in 2023. The roles are changing. If you’re job hunting, AI skills aren’t just a bonus anymore—they’re becoming standard. LinkedIn data shows that workers are now more than twice as likely to add AI skills to their profiles than they were six years ago. Take note and follow their example.

3. The USMCA Review Is Coming. North American Hiring Hangs in the Balance.

Recap: In July 2026, the United States, Mexico, and Canada will begin the first formal review of the USMCA trade agreement. Under Article 34.7, the three countries can extend the agreement for 16 years, initiate annual reviews, or begin a termination process. The review is expected to cover rules of origin, labor standards, nearshoring incentives, and broader issues, including migration and drug policy.

This story isn’t getting enough attention, but it matters more than most of the stories covered today.

The USMCA governs a $1.93 trillion trade corridor across three countries and 500 million people. The July review will determine whether the agreement gets extended to 2042, creating 16 years of regulatory predictability, or enters a period of annual reviews that could freeze cross-border investment decisions for a decade.

If you’re building distributed teams with talent from Latin America, this isn’t just background noise. Mexico brought in $40.9 billion in foreign direct investment in the first three quarters of 2025, up 14.5% from the year before. New investments jumped 218%. That growth is closely linked to the stability the USMCA brings. If the review goes well, it keeps the conditions that make nearshoring possible. Otherwise, the uncertainty might slow things down for years.

There’s some real irony here. In the same week a US tech company lays off 30,000 people to invest in AI, the trade rules that let companies hire skilled engineers in Latin America are being reviewed. Companies building strong teams now are watching both issues. While layoff headlines get attention, the trade review will shape what happens next.

Advice:

If you’re hiring or planning to hire in Latin America, mark July 2026 on your calendar. The results will affect everything from compliance costs to how fast you can grow a cross-border team. Don’t wait for the news to catch up.

That’s it for this week.

I began by saying the real headlines are stranger than any April Fools’ joke. There’s a bigger point: the labor market is changing shape. Jobs are being created as others are cut. Money is moving. Trade agreements are under review. Everything is shifting.

Successful companies won’t panic or ignore change. Instead, they’ll act thoughtfully, hire the right people, treat them well, and build teams they want to keep.

If you’re wondering what that could look like for your company, we’re always willing to chat at lupahire.com.

Until next time,
Joseph Burns
CEO & Founder, Lupa

Hey there,

This week, I keep wondering: Who actually benefits from the constant narrative that AI is taking everyone’s jobs in business media?

Think about it. Every CEO who blames AI when announcing layoffs gets a stock bump and avoids tougher conversations about overhiring or margin pressure. AI companies that stay in the “disruption” headlines benefit from the sense that change is inevitable. If everyone believes the transition can’t be stopped, the question shifts from whether to buy the tools to which ones to buy. Boards that want to freeze hiring now have an easy justification that sounds strategic instead of reactive.

That doesn’t mean AI displacement isn’t real. It is. Some jobs are being automated, entry-level hiring is slowing, and disruption is speeding up in certain areas. But “certain areas” is an important detail, and the main story doesn’t leave much room for specifics. Panic makes a better headline than nuance.

So this week I looked for the real numbers. Here are three stories, and none of them is about layoffs.

Let’s get into it.

🌐 News Shortlist

1. OpenAI Is Nearly Doubling Its Headcount

Recap: The Financial Times reported this week that OpenAI plans to grow its workforce from about 4,500 to around 8,000 employees by the end of 2026. The hiring covers product, engineering, research, enterprise sales, and a new role called “technical ambassadorship.” This function is meant to help large organizations actually use AI tools in their businesses. OpenAI is also expanding office space in several markets. The company is shifting from being mainly a research lab to a full-scale enterprise software company, now competing for the same B2B contracts as Salesforce, Microsoft, and Oracle.

The company most frequently associated with replacing jobs is now leading one of the biggest hiring waves in tech. That’s worth thinking about.

OpenAI’s growth shows where AI is in its business journey. Building models is just one part. Selling them to companies, integrating them into workflows, providing technical support, and renewing contracts all require different kinds of teams.

The “technical ambassadorship” category is the most interesting signal. It describes people who sit between the product and the customer — fluent enough in the technology to demo and troubleshoot, skilled enough in business to understand what the client actually needs and why adoption isn’t happening. This is a role that didn’t exist five years ago. It is in demand now because the gap between “AI can do this” and “your team is actually using AI to do this” is enormous, and closing it requires humans.

The bigger picture is that for every company cutting jobs because of AI, there are other companies—AI vendors, their partners, and the enterprise software teams built around these tools—that are hiring. These jobs are different from the ones being cut. They usually need more experience, technical skills, and strong communication. But these jobs do exist.

Advice:

If you’re hiring in enterprise software, AI infrastructure, or roles connecting technical products with business buyers, clarify what makes your job compelling and move quickly. Streamline your hiring process to remain competitive with OpenAI for top candidates.

2. Brazil Is Leading the Americas in Tech Hiring Confidence

Recap: ManpowerGroup’s Q2 2026 Tech Talent Outlook, released March 19, found that US employers reported a Net Employment Outlook of 41% for tech hiring — an 8-point increase from the prior quarter. Globally, tech hiring intent reached 45%, up 9 points year over year. Within the Americas, Brazil led at 63%, followed by Panama at 61% and Canada at 45%. India led globally at 69%. Colombia came in at 13%, reflecting more cautious employer sentiment. The survey was conducted from January 1 to February 3, before the outbreak of conflict in the Middle East.

This story isn’t getting much attention because layoff headlines are louder and easier to write.

An eight-point jump in one quarter is a real change in the US. It doesn’t mean tech hiring is back to 2021 levels—it’s not. But it shows that employers who are ready to grow are starting to hire. The jobs they’re filling are different now: more senior, more specialized, and more focused on AI. There are fewer open roles overall, but the ones that do open are filled more intentionally.

Brazil’s 63% is the number to watch. This isn’t a market under pressure—it’s growing. The local fintech and e-commerce scene, with Nubank reaching 131 million customers and Mercado Libre seeing 45% revenue growth, is creating strong demand for senior engineering, product, and data talent. Now, US companies hiring remotely are competing for the same people, which is pushing up pay for top candidates and making hiring harder than it was a year ago.

Colombia’s 13% tells a different story. It’s not about poor talent—the country has great professionals in tech, finance, and operations. Instead, it shows that local employers are cautious during a presidential transition year. For companies willing to put in the effort, such hesitation creates an opportunity: the best candidates are often more available and open to new roles when local demand is low.

Advice:

If hiring technical talent in Latin America for Q2 or Q3, expect tough competition in Brazil. Start searches earlier and highlight unique role benefits. In Colombia and other cautious markets, consider targeting roles that don’t require top-tier Brazilian expertise to ensure a smoother hiring process.

3. Stanford Economists Checked What AI Is Actually Doing to Jobs

Recap: At the 2026 Stanford SIEPR Economic Summit, a panel of economists—including Erika McEntarfer, former head of the Bureau of Labor Statistics—shared early findings on AI’s impact on hiring. Entry-level software developer hiring for ages 22–26 is down 20%. Call center hiring is down 15%. These are real declines, not just predictions. Still, the panel found no sign of a broad labor market collapse: unemployment for workers exposed to AI hasn’t gone up much, the overall job market is steady, and jobs with low AI exposure—like home health aides, skilled trades, and manual labor—are still growing. The researchers also noted a shift in professional fields like law, banking, and medicine, moving from pyramid-shaped hiring (many junior roles, fewer senior ones) to something more like a diamond or inverted shape.

The headline says “AI is eliminating jobs.” The more accurate story is that AI is cutting out the lowest level of certain knowledge jobs, but the middle and senior roles are mostly staying the same.

A 20% drop in entry-level software developer hiring doesn’t mean all software developers are being replaced. It means junior developers who used to write boilerplate, fix simple bugs, and handle documentation aren’t being hired because AI tools can do that work well enough. Senior developers who design systems and make key decisions are still being hired, often at higher pay. The same pattern is happening in law, finance, and operations.

The shift from a pyramid to a diamond shape in hiring is the most important finding. Junior roles serve two purposes: handling work that doesn’t need senior judgment, and developing future senior professionals. If AI takes over the first job, it becomes harder to do the second. Companies cutting entry-level jobs now are creating a talent pipeline problem that will show up in two or three years—and will be costly to fix.

For Latin American professionals, this data makes things clearer. The biggest interference is in entry-level, highly standardized technical jobs based in the US. The roles that are stable—or even growing—are the ones international hiring usually fills: experienced professionals with specific skills who can work independently and make good decisions. The talent pool I work with is mostly mid-to-senior level. That’s not by accident. It’s why demand has stayed strong.

Advice:

If you’ve cut or frozen entry-level jobs for AI efficiency, ensure you are still developing talent and mentoring staff. Don’t just automate tasks—actively build career growth programs or risk future skill gaps.

That’s it for this week.

Three different data points, but the same pattern: disruption is real and focused, and growth is also real and focused. The space between those two is where a smart hiring strategy happens right now. That’s what I help clients with. If you want to talk about it for your team, you can find me at lupahire.com or just reply here.

Until next time,
Joseph Burns
CEO & Founder, Lupa