150,000 Layoffs, Half Explicitly Blamed on AI: The Attribution Wall Has Broken
TexTak places [white-collar-displacement] at 70%, up from 67%, and today's news is the strongest direct evidence we've seen for that move. Over 150,000 tech jobs cut since January 2026. Nearly half — roughly 37,638 positions in Q1 alone — officially attributed to AI automation. Oracle. Amazon. Snap. Block. These aren't anonymous survey respondents hedging in a McKinsey report. These are public companies telling investors, in earnings calls and press releases, that AI is why the headcount is down. The attribution wall, which was the central barrier in our original thesis, has cracked open.
Our 70% reflects one core bet: that investor pressure for AI ROI would eventually force companies to publicly claim the productivity gains they'd been quietly harvesting. The thesis was never that AI displacement wasn't happening — it clearly was, through attrition and reduced hiring. The question was whether companies would say so out loud, given the PR risk of being seen as an automation villain. What moved us from 67% to 70% isn't just the volume of layoffs — 150,000 is a big number, but tech layoffs have been large before. What moved us is the attribution behavior. Jack Dorsey saying Block can operate with far fewer people because of AI. Snap disclosing that AI generates more than 65% of its new code and cutting 16% of staff in the same breath. That's the variable we were watching, and it's now observable at scale.
The strongest counterargument isn't that this is overstated — it's that we're measuring the wrong thing. Our forecast asks whether a 'major layoff wave' will be 'explicitly attributed to AI automation.' Oracle, Amazon, and Snap satisfy that criterion individually, and collectively they satisfy it overwhelmingly. But a skeptic could reasonably argue that tech-sector layoffs are a special case: these companies have sophisticated IR teams who understand that 'AI efficiency' is a better story for investors than 'cost-cutting,' so attribution may reflect narrative management as much as operational reality. In other words, companies may be over-attributing to AI precisely because it flatters them. That's a real critique, and it means our evidence is slightly softer than it looks — the attribution is real, but its causal accuracy is less certain.
The Stanford HAI finding compounds this picture in a way we didn't fully anticipate: employment among software developers aged 22–25 fell nearly 20% since 2022. That's not a layoff announcement — that's a structural labor market signal, and it's the kind of evidence that's harder to explain away as narrative management. It's also the demographic most exposed to AI coding tools, which is exactly where our thesis predicted displacement would concentrate first. This is proximate evidence, not direct proof of our forecast target, but it's consistent with the thesis in the right demographic segment at the right time.
What would move us back below 60%? If Q2 earnings calls show a reversal — companies attributing headcount recovery to AI-driven hiring in new roles that offset the cuts — that would challenge the 'wave' framing. If major companies begin walking back AI attribution under political pressure (the 'automation villain' narrative gains traction in an election cycle), we'd revisit. We're also watching whether the 150,000 figure represents a concentrated burst or the leading edge of a sustained trend. If layoff pace decelerates sharply in Q2 with no new major public attributions, the wave may be more episodic than structural. We'll know more by June.