When 'AI Layoffs' Aren't Really About AI

"We're using more AI, so we need fewer people." It is a clean story, suitable for an earnings call and a press release. Increasingly, it is a story that the underlying data does not support, and analysts have started calling the practice "AI washing."

An analysis covered by Moneywise looked at the 80,000 tech jobs cut in the first quarter of 2026 and found the AI rationale was being applied broadly, including to roles in functions where AI has yet to demonstrate the productivity gains executives describe. A Department of Labor Chief Innovation Officer told reporters bluntly that some firms have been "scapegoating AI."

What is actually driving the cuts

Several pressures explain most of the recent tech layoffs.

The first is the bill from pandemic-era hiring. Many of the firms cutting hardest in 2026 doubled headcount between 2020 and 2022, and have spent the last eighteen months working back to a sustainable run rate. The Federal Reserve Bank of Dallas has said it considers wide AI-driven job loss unlikely within the next decade.

The second is interest rates. Capital was effectively free during the zero-rate years and then suddenly was not, which forced a hard reset on the cost structure of growth-stage companies.

The third is the cost of AI itself. Amazon, Meta, Google and Microsoft are expected to spend a combined $650 billion on AI infrastructure in a single year. That money has to come from somewhere, and payroll is one of the largest controllable lines in a corporate budget. As Forrester analysts put it to Tom's Hardware, AI is paying for AI.

Where AI really is reshaping work

None of this is an argument that AI is having no effect on the labour market. It clearly is. Hiring for entry-level roles in AI-exposed occupations appears to be slowing, with the earliest visible disruption showing up as fewer first jobs rather than waves of replaced incumbents. Specific functions, sales representatives among them, did see a measurable headcount decline between mid-2024 and mid-2025.

For more on how this plays out for software engineers in particular, see our earlier piece on whether AI is replacing programmers.

The honest version of the story is messier than the press-release one. AI is a real pressure on certain roles. It is also a convenient label for cuts that were always coming.