55% of 2026 Layoffs Cite AI Explicitly. The Attribution Firewall Has Broken.
textak places the probability of a major AI-attributed layoff wave at 73%, up from 72% last month. For months, our core thesis has been that the phenomenon was real but companies would avoid public attribution for PR reasons. Today's data from Skill Syncer breaks that assumption open: 55% of layoff events in the first half of 2026 explicitly cite AI, automation, or machine learning as a driving factor, affecting 152,415 workers across 135 companies. That's not quiet displacement. That's public declaration.
Our 73% reflects two reinforcing dynamics: the underlying displacement is accelerating, and the attribution behavior we said would lag the phenomenon has now arrived simultaneously. Let's be precise about what the Skill Syncer data actually proves. This is direct evidence that companies are publicly attributing layoffs to AI — which is the specific behavioral threshold our forecast targets, not just the displacement itself. Oracle's 30,000-person reduction is the largest single event, but the 247 total layoff events spread across tech, finance, and healthcare tells the more important story: this isn't one dramatic headline, it's structural and sector-wide.
Why did the attribution firewall break? Our working hypothesis is that the 56% AI skills wage premium reported by Digital Applied and the Stanford AI Index changed the internal narrative. When companies can simultaneously announce layoffs and hiring for AI-skilled roles at premium wages, the PR calculus shifts — attributing to AI signals strategic transformation rather than cost-cutting desperation. The 3.2:1 demand-to-supply ratio for AI skills gives executives a framing tool that makes public attribution less risky than we modeled 18 months ago.
The strongest counterargument to full forecast resolution is definitional. Our forecast requires a 'layoff wave explicitly attributed to AI automation' — and a reasonable challenger could argue that 55% citation rates across smaller events still doesn't constitute a single watershed moment with clean causal attribution, as opposed to AI being one factor among several cited in restructuring announcements. We've considered this. The 135-company breadth and the explicit, not euphemistic, language across the filings convinces us we're inside the resolution zone rather than approaching it. But we acknowledge this is partly a judgment call on language standards, not a clean binary.
What would move us below 60%? If Q3 earnings cycles show companies reverting to euphemistic 'efficiency' language as the PR environment shifts, or if a major employment discrimination lawsuit tied to AI layoff attribution creates legal incentive to stop citing AI explicitly — that would be a genuine signal. We're watching Q3 earnings call transcripts specifically for the ratio of 'AI-driven efficiency' to 'workforce optimization without AI attribution.' Until then, 73% holds, and honestly, the pressure is upward.