The Attribution Wall Has Broken: AI Displacement Is Now a Public Fact
textak has held a 73% probability that a major, explicitly AI-attributed layoff wave would materialize — and as of mid-June 2026, the data has arrived in force. SkillSyncer's tracking shows 55% of all June layoff announcements now explicitly cite AI as the driver, affecting 152,415 workers across 135 companies. Oracle's 30,000-person cut is the single largest of the year. The forecast variable we always said mattered most — not whether displacement was happening, but whether companies would publicly attribute it — has now resolved in the affirmative at scale.
Our 73% reflected a specific bet: that investor pressure for AI ROI would eventually outweigh the PR risk of explicit attribution. The historical pattern was attrition-based displacement with vague 'restructuring' language. What we're seeing now is categorically different. Goldman Sachs putting a number on it — roughly 11,000 net AI-attributed losses per month — gives institutional cover for other firms to use the same framing. Once Goldman is on record, the attribution risk collapses. This is the mechanism we were watching.
The honest counterargument here was always that companies would continue to absorb AI displacement quietly through attrition, never triggering the kind of concentrated, attributable wave the forecast required. That counterargument was reasonable — and it held longer than we initially expected. What changed isn't the scale of displacement, which has been building for two years. What changed is the corporate communication calculus. When 25% of all layoff announcements in March cited AI as the leading driver, the reputational math flipped: being seen as 'not using AI' became more costly than being seen as displacing workers with it.
The law firm data reinforces the structural dimension. AI contract review tools reducing junior associate hours by 40% represent exactly the kind of quiet, incremental displacement that never shows up in headline layoff numbers — but reshapes staffing pyramids over 18-24 months. The SkillSyncer data captures announced cuts; it almost certainly undercounts the attrition-based layer happening simultaneously in legal, finance, and back-office functions.
What would move us below 65%? If Q3 layoff data shows the June spike was a seasonal anomaly — if AI attribution drops back below 20% of announcements by September — we'd revisit the durability thesis. But we're not watching for that. What we're watching now is whether this explicit attribution pattern triggers the political response that makes companies walk it back. If Congressional hearings or administration pressure creates new PR risk for AI attribution language, companies could revert to euphemism. That's the scenario that complicates the next chapter of this forecast.