The Attribution Threshold Has Been Crossed — But Our Forecast Definition Hadn't Kept Up With It

Here's the problem we created for ourselves. The published forecast reads 'first major layoff wave explicitly attributed to AI automation' — full stop. No qualifier about what medium that attribution must appear in. But the actual resolution criteria we've been applying internally requires explicit attribution in SEC-filed earnings communications or formal press releases, not forum remarks, blog posts, or CEO speaking appearances. That distinction is doing significant work in holding this at 73%, and readers cannot independently evaluate it because it wasn't in the published definition. We're fixing that now: the forecast resolves YES when a major employer explicitly attributes a reduction-in-force of 1,000+ roles to AI automation in a Form 8-K, 10-Q, or formal press release filed with or distributed through SEC channels. GitLab's Act 2 restructuring announcement, in which the company stated AI agents are 'central to its operational transformation' and linked cost savings to AI product reinvestment, comes closest to meeting this bar — it was a formal announcement, not a forum comment. We are reviewing whether it crosses the threshold. If it does, this forecast resolves, not moves to 74%.

On the evidence quality: the 150,000+ figure from programs.com is a third-party tally of job losses where AI was cited as a contributing factor across sources of varying formality. Some of those citations are executive statements at investor forums. Some are internal memos reported by journalists. Some may approach SEC-level attribution. The number is real and directionally meaningful — it's circumstantial evidence that the phenomenon is happening at scale, and it's consistent with our thesis. But it is not a direct count of formally attributed AI displacement in the specific sense our resolution criteria require. We should have said that in the lede, not buried a caveat three paragraphs deep.

The Gartner finding — 80% of companies that piloted AI reported workforce reductions — is useful context but it is not direct evidence for public attribution behavior. Gartner measured self-reported reductions in a survey of pilot participants. A company telling a Gartner researcher 'yes, we cut roles during our AI pilot' is a meaningfully different act from that same company filing an 8-K saying 'we are eliminating 1,200 positions due to AI-driven operational efficiency.' The first is private survey disclosure; the second is the public attribution event our forecast is tracking. We're using the Gartner data as a proximate signal — it tells us the underlying phenomenon is widespread, not that the public attribution threshold has been crossed.

The counterargument that honestly keeps us up at night isn't the attrition-versus-layoff distinction we've emphasized before. It's this: Jamie Dimon and GitLab's leadership may have strategic incentives to attribute cuts to AI that are independent of AI being the actual driver. In a tightening revenue environment, 'AI-driven transformation' is a more favorable narrative than 'cost reduction due to margin pressure.' If AI attribution is functioning as convenient cover for cyclical or competitive restructuring, then what we're observing is a communications artifact — companies learning that AI framing is investor-friendly — rather than genuine evidence that AI is the actual displacement mechanism. We don't have a clean answer to this. What would partially address it: checking whether the companies explicitly naming AI also show increased AI-related CapEx and reduced headcount in roles that AI demonstrably performs. GitLab redirecting cost savings toward AI product investment is one data point in the right direction. We're watching for more.

What moves this above 80%: a Fortune 500 employer files an 8-K or issues a formal press release explicitly attributing a 1,000+ person reduction-in-force to AI automation, with the word 'AI' or 'artificial intelligence' appearing in the formal document — not in an executive's conference remarks. What drops us below 55%: if the next six months of earnings season shows companies consistently avoiding AI attribution language in formal filings even while their executives discuss it publicly, suggesting the formal attribution behavior we're forecasting isn't coming regardless of the underlying phenomenon's reality.