AI is now part of the layoff language
The latest layoff data makes it harder to treat AI as a distant worry. In its May 2026 report, Challenger, Gray & Christmas said U.S.-based employers announced 97,006 job cuts in May, the highest May total since 2020. The firm also said artificial intelligence led all stated reasons for job cuts for the third month in a row, with 38,579 cuts attributed to AI in May, or 40% of all announced cuts that month. For the year through May, AI had been cited in 87,714 cuts, already above the 54,836 cuts attributed to AI in all of 2025. :contentReference[oaicite:0]{index=0}
That does not prove AI directly performed every job that was cut. Layoff explanations are not laboratory evidence. Companies may cite AI when they are also dealing with restructuring, investor pressure, weak demand, mergers, bankruptcies, or cost controls. Challenger itself noted that companies are “restructuring aggressively” as they reposition for an AI-driven economy. :contentReference[oaicite:1]{index=1}
Still, the wording matters. Once AI becomes a common layoff reason, it gives companies a cleaner story for messy workforce decisions. A department can shrink because leaders believe fewer people will be needed later. A hiring plan can be paused because automation might cover part of the workload. A junior opening can be redesigned into a broader role because managers assume AI will handle the basics. The worker experiences all of this as instability, even when no one can point to one tool that “took” the job.
The bigger shift is happening at the bottom of the ladder
For job seekers, the scariest part may be what is happening to entry-level work. PwC’s 2026 Global AI Jobs Barometer, which analyzed more than one billion job ads across six continents, found that AI is increasing employer demand for human skills such as judgment, creativity, and leadership. In U.S. entry-level roles most exposed to AI, PwC found those jobs are now seven times more likely to require traditionally senior-level skills. Those AI-exposed entry-level roles grew 35% since 2019, while other entry-level roles declined by 10%. :contentReference[oaicite:2]{index=2}
That sounds positive at first. More advanced entry-level roles could mean better work, faster development, and higher pay for people who can get hired. But it also changes who gets access. A normal early-career role used to offer room to be new: draft the report, clean the data, schedule the meeting, summarize the notes, prepare the first version, learn by watching someone more experienced revise it. If AI can now produce the first draft, organize the spreadsheet, or summarize the call, companies may ask why they need as many people doing the early-pass work.
The problem is that those early-pass tasks were not just low-value busywork. They were training. They taught workers what good judgment looks like before they were expected to show it alone. When companies automate the beginner layer, they may still need judgment, taste, client awareness, risk awareness, and communication. But they may expect applicants to show those skills before they have had a realistic place to build them.
“Entry-level” no longer always means beginner
This is why so many job postings feel off to applicants. A role may be labeled associate, coordinator, analyst, or junior specialist, but the description asks for ownership, stakeholder management, strategic thinking, executive communication, and comfort with AI tools. None of those requirements is automatically unreasonable. The issue is the bundle. When an entry-level title carries mid-level accountability, the risk shifts from employer to applicant.
For a recent graduate or career changer, that means the application process becomes less about proving potential and more about proving you can already operate. The person who did internships, freelance work, student consulting, military service, caregiving logistics, or a portfolio project may be able to tell that story. The person who simply needs a first professional chance may be filtered out before anyone sees their ability to learn.
This also helps explain why the job market can feel contradictory. There may still be postings. Companies may still say they are hiring. Some AI-exposed companies may even grow faster than peers. But the available jobs may not match the workers trying to enter the market. A ladder can exist on paper while still being too high off the ground.
Mid-career workers are not just being “freed up”
The other side of the story is what happens to the workers who remain. AI tools are often sold inside companies as a way to remove drudgery and create time for higher-value work. Sometimes they do. But recent data suggests workers are also absorbing a new layer of supervision.
Glean’s Work AI Index 2026, from the Work AI Institute, surveyed 6,000 full-time digital workers across the United States, the United Kingdom, and Australia. It found that 87% of digital workers use AI at work and 75% say it makes them more productive, but only 13% say their organization is performing significantly better because of it. The report says workers spend an average of 6.4 hours per week “botsitting,” meaning they are feeding AI missing context, checking outputs, debugging mistakes, rerunning prompts, and cleaning up answers that sound confident but are wrong. :contentReference[oaicite:3]{index=3}
That is not a small side task. It is most of a workday. And it often falls on people who are already carrying the pressure of smaller teams, tighter budgets, and vague productivity goals. A manager may ask why a project is not moving faster now that AI is available. The employee may be spending the saved time verifying what the AI produced, rewriting it into something usable, or explaining why the output cannot be trusted.
This is the part of AI adoption that does not always appear in headcount plans. A company may reduce junior support because the tool can produce a draft. Then a mid-career worker has to prompt the tool, correct the draft, check the facts, adjust the tone, protect the client relationship, and take responsibility for the final result. The task did not vanish. It moved up the ladder.
What job seekers should watch in postings
For job seekers, the practical move is not to panic or pretend AI does not matter. It is to read job descriptions more carefully. Watch for junior titles that ask for broad ownership without clear training. Watch for “AI-first” language that does not explain the actual workflow. Watch for roles that combine strategy, execution, reporting, client communication, automation, and quality control under one entry-level salary band.
Applicants should also change how they describe their own value. Basic AI familiarity is becoming less distinctive. Better signals include examples of judgment: how you checked an AI output, improved a messy process, caught an error, explained a tradeoff, turned unclear information into a decision, or used a tool without letting the tool own the work. Employers asking for senior-like skills are really asking, fairly or not, whether you can be trusted with ambiguity.
What workers should watch inside companies
For employed workers, the warning sign is not just “we are rolling out AI.” The warning sign is AI adoption without workflow design. If leaders add tools but do not change deadlines, staffing, review standards, data access, training, or accountability, the missing structure will usually be supplied by workers’ evenings, attention, and stress.
It is reasonable to ask practical questions: Who checks AI-generated work? What kinds of work should not use AI? How is saved time measured? Are employees being evaluated on output volume or quality? Is AI supervision part of the job description, or is it invisible labor? Those questions matter because untracked work has a habit of becoming unpaid expectation.
The new career risk is a thinner path
The AI labor story is not just a replacement story. It is a redesign story. Some jobs will be cut. Some will grow. Some will become more interesting. Some will become more exhausting. But the most important shift for many white-collar workers may be the shrinking space between “no experience” and “already trusted to decide.”
That space is where careers used to form. It is where people learned how to write the client email, read the room, question the data, recover from a mistake, and understand what good work looks like. If companies remove too much of that layer, they may save money in the short term while creating a workforce that is expected to have judgment no one invested in developing.
AI may not take your job in one clean moment. The more realistic risk is that it changes the job above you, removes the job below you, and leaves everyone competing for roles that require more proof, more judgment, and less training than before.
Leave a Reply