Is AI Taking Your Job? Part 2: CEOs Blame the Fancy Computers, But The Data Is Murky

Tech layoffs hit their highest single month in years in May, and roughly 120,000 tech roles have now been cut in 2026. Is that the real reason, or the most convenient one?

Part 1 looked at whether the communities that host data centers actually see job growth from them. The answer there, based on Brookings research, was a qualified yes, but the fancy computers themselves barely employ anyone.

So, do the companies pouring hundreds of billions into AI grow their own workforces because of it, or is AI just the label attached to cuts that were coming anyway? Is AI really behind the 2026 Tech Layoffs? Here, the evidence pulls in three different directions at once, and none of them fully resolves.

The Case for "No"

The top tech titans have all cut jobs while investing heavily in AI: Amazon eliminated 30,000 corporate roles in three months as Jassy explicitly tied the cuts to AI-driven efficiency gains; Meta laid off 8,000 while shifting 7,000 more into AI-focused roles; Microsoft's headcount kept declining through voluntary buyouts before it cut another 4,800 roles in July, insisting the jobs weren't AI-replaced even as it touted AI's effect on how work gets done; and Google trimmed a third of its small-team managers and an estimated 1,500 to 3,000+ engineers without ever announcing a single formal number. The pattern repeats down the rest of the industry — Oracle (21,000, per its own SEC filing), PayPal (4,500+), IBM (15,000+ since 2024), Intuit, Cisco, Cloudflare, Coinbase, Snap, Atlassian, Dell, Block, Salesforce, and GitLab have all announced cuts this year with AI cited as a factor, frequently in the same breath as record revenue.

Gartner surveyed 350 executives at companies piloting AI and found roughly 80% had cut jobs with no correlation between those cuts and improved ROI. Of 23 S&P 500 companies that tied layoffs to AI, 56% have traded down since the announcement, by an average of 25%. The money doesn't always land somewhere better, either. Teradata's CEO told employees not to expect a raise, writing that the company would fund its AI investment by reallocating the 2026 salary-adjustment budget. In other words, the line item for compensating workers is being poured into the fancy computers. And the scale often doesn't add up. Meta's roughly 8,000-person layoff saves an estimated $2.4 billion a year against AI capital spending projected as high as $145 billion in 2026. Morgan Stanley estimates data-center construction costs alone will hit nearly $2.9 trillion industry-wide from 2025 through 2028. Payroll cuts aren't funding this buildout. The gap is too large.

The Case for "Yes"

A June 2026 study by Ramp and Revelio Labs linked real AI vendor spending data to workforce records across more than 21,000 U.S. firms. Companies investing most heavily in AI grew headcount by about 10% over the following two years, with entry-level hiring up 1.15%.

The researchers are careful about what this does and doesn't show. It's correlation, not causation. Heavy AI adopters were already larger, faster-growing, and more likely to be venture-backed before they adopted anything. The kind of company that spends heavily on AI is also the kind of company that is already growing. Those might be the same companies for reasons that have nothing to do with AI. Still, it's real data pointing the opposite direction from the layoff headlines.

The Case for "We Can't Trust Either Side's Framing"

There's a third complication that undercuts both arguments above. Companies may not be telling the truth about why they're cutting jobs in the first place.

MIT economist David Autor told NBC News: it's much easier for a company to say it's cutting jobs for "AI-related efficiencies" than to admit it overhired, overspent, or is bracing for a weaker economy. Whether or not AI is the real reason, he noted, a company is often better off letting AI take the blame, or the credit.

MIT's Paul Osterman calls this "AI washing," using the trendy technology of the moment as a convenient cover for layoffs. Cisco's stock jumped 13% after it announced AI-linked layoffs, which tells you the market rewards this framing regardless of whether AI is doing anything at all. Even OpenAI's Sam Altman has pushed back publicly on the trend, arguing that the recent layoffs are driven more by pandemic-era overhiring, slower growth, and margin pressure than by AI.

This means at least some of the data feeding both "AI is destroying jobs" and "AI isn't destroying jobs" narratives may be built on inaccurate company self-reporting. Nobody outside the company knows for certain which layoffs are true AI substitution and which are AI-flavored PR.

So What Do We Actually Know?

Less than the headlines suggest. Three things can be true simultaneously. 

Some companies are cutting jobs and AI is honestly to blame. Some are cutting jobs and blaming AI dishonestly, and some are genuinely growing their workforce alongside heavy AI investment for reasons that may or may not have anything to do with AI itself. The data available at this time does not paint a clear picture.

For workers, this means tracking where the real hiring is happening. The Ramp/Revelio gains were concentrated almost entirely in high-intensity AI adopters, in specific sectors (information, finance, professional services), and even there, mostly among companies that were already fast-growing before they touched AI. If you're job-hunting or thinking about a pivot, that's a more useful signal than any single company's stated reason for a layoff.

For investors, this means treating "AI-related layoffs" and "AI investment" as two separate bets that need two separate kinds of evidence. A stock bump on layoff news doesn't confirm AI is working. The Ramp/Revelio 10% headcount growth is a genuine positive signal, but it's a correlation among companies that were unusual to begin with, not proof that the AI spending itself is what's paying off. Following the ROI data, not the press release, is the only way to tell which of those is happening at any given company.

The murkiness itself is the finding. Nobody - not the researchers, not the CEOs, not the market - currently has a clean answer to whether AI investment creates jobs. Anyone telling you otherwise, in either direction, is telling you more than the data currently supports.

Written with AI assistance; edited and fact-checked by the author.

Photo by Jakub Żerdzicki on Unsplash