56% of Layoffs Now Cite AI Explicitly. The Attribution Wall Has Broken.
textak has held a 73% probability that a major AI-attribution layoff wave would become publicly undeniable — and today's evidence is about as direct as we've seen. SkillSyncer's analysis of 267 layoff events through June 27, 2026 shows 56% explicitly naming AI, automation, or machine learning as a contributing cause, affecting 156,270 workers. Jamie Dimon confirmed in February that JPMorgan has already experienced AI-driven displacement. The forecast is not about whether displacement is happening — it's specifically about whether companies would publicly attribute it. The answer, increasingly, is yes.
Our 73% reflects three things we've been watching: back-office headcount reduction, reduced junior hiring in technical roles, and — most critically — whether corporate leadership would move from quiet attrition to explicit public attribution. The third factor was always the hardest to forecast, because the PR logic cuts against admission. For the first 18 months of this cycle, companies systematically avoided the word 'AI' in layoff communications, preferring 'restructuring' or 'efficiency.' That pattern has now visibly broken.
The SkillSyncer data is direct evidence, not circumstantial. This isn't 'companies are buying AI tools' or 'AI capabilities could displace these roles' — it's companies explicitly citing AI in documented layoff events at a 56% rate across 267 discrete instances. The BCG analysis adding that 50-55% of US jobs will be 'substantially reshaped' in two to three years is proximate evidence — it measures conditions forming, not the attribution behavior itself. But Dimon's February statement is particularly significant because it comes from the CEO of the largest US bank, in a recorded public forum, with explicit language about displacement already underway. That's the kind of attribution event our forecast was specifically tracking.
Honestly, the part of this thesis that still keeps us up at night is the distinction between 'attribution in layoff communications' and 'attribution in public forums.' SkillSyncer's methodology — analyzing 267 events — is real, but we haven't independently verified their classification criteria for 'explicitly cites AI.' If their bar is loose (any mention of automation in any company communication counts), the 56% figure overstates public attribution in the specific sense we care about. There's also a genuine question about survivor bias: are we measuring the subset of companies willing to disclose, while a larger universe of AI-driven attrition goes unannounced?
What would move us above 80%: a Fortune 100 company filing a 10-K with a dedicated 'AI-driven headcount reduction' line item, or a congressional hearing where multiple CEOs testify to AI attribution explicitly. What would drop us below 60%: evidence that the 56% figure reflects loose classification methodology, or a coordinated shift back toward 'restructuring' language as political blowback intensifies. Neither of those has happened. We're holding 73% — and if anything, the Dimon disclosure and SkillSyncer volume together make us think the number has more upside risk than downside from here.