Is AI really behind the layoffs? What the data shows

AI layoffs dominate the workforce headlines in 2026, but the story behind them is messier than a single number can capture. Layoff trackers cite AI as a leading reason for job cuts, executives reference “AI transformation” in earnings calls, and employees increasingly feel like they’re competing against a system that never clocks out.

Labor market data tells a more complicated story. AI is measurably shrinking the opportunity in a handful of specific, well-defined occupations, yet the broader hiring numbers show no clear evidence that AI adoption is behind the overall slowdown in the job market. Both are true, and the gap between them is exactly where employers get their communication wrong.

Today, AI seems to be replacing a narrower set of jobs than the headlines suggest. The effect is concentrated by task and career stage rather than spread evenly across the workforce.

Researchers at the Federal Reserve Bank of New York, using an occupational exposure framework built by Anthropic economists, found that the roles most exposed to generative AI include computer programmers, customer service representatives, and data entry specialists—jobs built around tasks that large language models now do reasonably well, like drafting, formatting, retrieving information, and answering routine questions.

The effect gets sharper by career stage. Early-career workers in AI-exposed occupations are losing ground relative to their peers, while more experienced workers in the same occupations are holding steady or continuing to grow. That divergence is the clearest reproduced finding in the research so far—a meaningfully different story than “AI is replacing jobs” as a blanket claim.

The data also helps determine if AI is being used to automate a task or advance the person doing it. Stanford’s research found that employment keeps growing in roles where employees use AI to research a topic, draft a first pass, or check their own work before it’s shipped. The decline is concentrated in roles where AI performs a task end to end, with no human step in between.

Software development is the clearest case study inside that data. According to the Stanford study, entry-level developer employment peaked during the hot labor market of 2022 and has been declining since, even as employment for developers over 30 continued to climb during the same period. The pattern holds for customer service and other roles built around well-defined, repeatable tasks: The jobs aren’t disappearing outright, but the on-ramp of entry-level roles into that work is narrowing.

Zoom out from specific occupations to the broader hiring numbers, and the “AI-is-gutting-jobs” narrative becomes more uncertain. A Federal Reserve staff analysis found no evidence so far that AI adoption is reducing firms’ overall job postings. If anything, firms and industries with higher AI adoption see a marginal increase in openings relative to lower-adoption peers, and the broader, post-pandemic slowdown in postings doesn’t appear to be driven by automation.

Moreover, the Federal Reserve Bank of New York found that firms overwhelmingly plan to bring AI into the workplace by retraining existing employees, rather than through broad reductions in hiring, and that few businesses report AI-driven layoffs directly, even as AI adoption continues to climb.

Ultimately, reallocation, not elimination, is the more accurate description of what’s happening overall: AI absorbs certain tasks, employers shift the freed-up hiring capacity toward other roles or skill sets, and the entry-level rungs of the most exposed occupations bear the disproportionate cost of that shift.

For workforce planning purposes, treating “AI is replacing jobs” as a settled fact will lead to different staffing decisions than as a task-level shift impacting specific occupations. The first framing justifies broad, reflexive headcount cuts. The second points toward something more precise: identifying which roles involve fully automatable tasks, redeploying or retraining the people in adjacent roles, and reserving workforce reductions for cases where the work itself, not just the narrative around it, has actually changed.

Need help building your workforce plan? Careerminds’ Basic Guide to Workforce Planning breaks down the process to create a more efficient, productive workforce structure. Click below to explore how to assess workforce gaps, increase retention, and segment your staff for success.

If the aggregate hiring data doesn’t show AI gutting the job market, why does AI keep showing up as the number one employer-cited reason for layoffs in 2026? It’s worth separating the reason companies give from the reason a layoff actually happened.

Challenger, Gray & Christmas, which tracks announced job cuts by employer-stated reason, has recorded AI as the leading cited cause of layoffs from March to July 2026. AI was cited for about 23% of all cuts announced in the first half of the year, with the technology sector absorbing close to a third of all 2026 job cut announcements. At the same time, overall layoff volume actually ran about 40% below the same period in 2025, and hiring plans in July ran roughly a quarter above where they stood a year earlier.

The firm’s workplace experts put it plainly: AI is shifting the labor market, but isn’t dismantling it. The combination of AI’s rising share of cited layoff reasons, overall layoffs trending down, and hiring picking back up is the clearest evidence that AI has become the go-to explanation employers reach for, whether or not it’s the most complete one.

For many companies, “AI” is a more palatable reason to give than the alternatives. Naming AI as the driver of a reduction sounds more forward-looking and strategic than a missed revenue target, over-hired team, or straightforward cost-cutting decision. In the end, some share of “AI layoffs” headlines reflect genuine automation shifts, but in many cases, just reflect a company reaching for the most defensible-sounding label available.

For HR leaders and CHROs, data eventually turns into a communication problem. Two mistakes involving layoff communication show up often for organizations navigating this kind of transition.

The first is, as we’ve seen, over-attributing a layoff to AI when the real driver is somewhere else. “AI transformation” sounds inevitable and depersonalizes a hard decision. Still, many employees can tell when a stated reason doesn’t match what they’re seeing, and a mismatch between the official story and reality erodes trust with both departing and remaining employees.

The second mistake is the opposite: under-communicating when AI genuinely is the reason, because the explanation feels awkward or exposes a strategic bet that didn’t fully pay off. Employees facing a genuinely AI-driven reduction deserve a specific explanation of what changed, not a vague reference to “operational efficiency” or “the future of work.”

Beyond the general framing, a few concrete practices separate honest AI-layoff communication from vague hand-waving:

• Name the task, not the technology: Say what specifically got automated (e.g., a support queue, a reporting process, a first draft of code review) rather than gesturing at AI as a whole.
• Give the actual numbers: If AI is one of several factors in a reduction, say roughly what share. Stating upfront that a layoff was “half AI and half a slower renewal quarter” will be more credible to employees.
• Don’t retrofit the reason after the fact: If the decision was made for performance or cost reasons, don’t relabel it as AI-driven because it sounds more forward-looking. Employees compare notes, and the mismatch will surface fast.
• Flag it early for roles at real risk: If a function is being restructured around AI, tell the team before the layoff. Early warning gives people time to reskill or look for other opportunities internally before they’re let go.
• Pair the explanation with a path forward: Say what skills or roles the organization is hiring for next, so the message isn’t just about jobs being eliminated but also about where the organization is headed and how employees can advance if they stay.
• Match support to who’s affected: Don’t default to a one-size-fits-all outplacement package. Early-career employees losing entry-level, AI-exposed roles need different coaching than a mid-career employee changing companies.

If you’re unsure how to communicate a layoff to employees, click below to download our free Careerminds Layoff Script, which provides five clear steps to guide you through the process with empathy and professionalism.

The honest answer to “is AI really behind the layoffs” is neither apocalyptic nor dismissive. The data doesn’t support the broad AI-based hiring collapse that some headlines describe. That said, AI has produced a real, measurable effect on specific occupations and workers, especially early-career employees whose roles are built around the exact tasks AI now handles.

For employers, the practical takeaway is to be as honest and precise as possible. Name the specific change when AI is genuinely the cause, be honest when it isn’t, and recognize that the employees most affected by this shift—disproportionately early-career workers—often need more targeted support to land their next role than a standard, one-size-fits-all program provides.

Staying ahead of AI-driven disruption also means staying ahead of AI itself. Organizations that track how technology is evolving and invest early in upskilling and reskilling their staff are better positioned than those that react only after roles become obsolete. Employees who build AI-adjacent skills are more likely to stay engaged and with the company, helping the organization retain its best people while making the most of its AI investment.

Careerminds helps HR leaders build layoff communication that holds up to scrutiny and outplacement support for today’s market, including the specific challenges AI-exposed and early-career employees are facing. Click below to speak with our experts about building a transition plan that’s honest about what’s actually driving your workforce changes.

AI is replacing specific tasks more than replacing jobs wholesale. Broader hiring data, including a Federal Reserve staff analysis that combines job-posting figures with Census Bureau survey data, show no clear evidence that AI adoption has reduced overall job postings; firms with higher AI adoption aren’t posting fewer openings than their peers. AI’s measurable, real effect is in specific, highly exposed occupations, especially for early-career workers whose employment in those fields has declined relative to peers in less-exposed roles.

AI is having the clearest effect on jobs built around routine, well-defined tasks. Computer programming, customer service, and data entry rank among the most exposed occupations, according to research by the Federal Reserve Bank of New York using an AI-exposure framework developed by Anthropic economists. The effect is heavily concentrated among early-career workers in these fields; more experienced workers in the same occupations have largely held steady or continued to grow.

AI is automating specific tasks within roles, such as drafting, formatting, retrieving information, and handling routine customer inquiries, rather than eliminating entire occupations outright. This shows up first at the entry level, where jobs are often built around these kinds of repeatable tasks, before employees develop the judgment and context that make a role harder to automate. Employers describe this in earnings calls and layoff announcements as automating roles, consolidating functions, or reallocating budget toward AI infrastructure.

Employers are automating tasks with AI to cut costs and speed up routine work, and increasingly citing AI as their reason for job cuts. AI has led Challenger, Gray & Christmas’ monthly tracker of employer-stated layoff reasons for five consecutive months through July 2026. That said, these reasons are self-reported rather than independently verified, and AI is often a more palatable explanation to publicize than alternatives like a missed revenue target or straightforward cost-cutting, so not every layoff that’s labeled “AI-driven” reflects the same degree of actual automation.