Anthropic model pegs Amodei's 17.9% jobless warning to its most extreme scenario

Anthropic has published an economic model that projects three possible paths for the US economy through 2030, and the most alarming of them is the one its own chief executive has been describing in public. In that scenario, economic output doubles roughly every 4.5 years while unemployment among knowledge workers climbs to 17.9 percent. The framing matters because Dario Amodei spent May warning that AI could wipe out large swaths of white-collar work — and the company's own numbers now place those warnings in the outlier bucket.

The model lays out a spread rather than a single forecast. At one end sits a slow-absorption path in which AI diffuses gradually and labor markets adjust without a sharp break. At the other sits the doubling scenario, where capability gains translate almost directly into displaced cognitive work. The 17.9 percent knowledge-worker unemployment figure is the headline number from that extreme case, and it is the only branch that matches the severity Amodei described in his public remarks.

That gap between the CEO's megaphone and the model's central estimates is the real story here. Anthropic is not walking back Amodei's comments, but the publication effectively tags them as a tail risk rather than a baseline. For a company that sells enterprise AI, the distinction is not academic: a model that says mass white-collar displacement is possible but unlikely gives customers and policymakers a very different planning horizon than one that treats it as the expected case.

The three-scenario structure also puts Anthropic in line with a broader shift in how AI labs talk about economic impact. Instead of a single dramatic prediction, the industry is increasingly publishing bounded ranges with explicit assumptions about adoption speed, task automation and how quickly firms reorganize around new tools. Those assumptions do most of the work, and they are exactly what outside economists will scrutinize. A 4.5-year output doubling is an extraordinary growth rate by historical standards, well outside anything the US has sustained, which is part of why the scenario reads as a stress test.

What to watch next is whether the model's intermediate scenario gets the same attention as its extreme one. Amodei has not retreated from his warnings, and Anthropic has not said the worst case is impossible — only that it is one branch among several. The open question is which assumptions would have to hold for the middle path to break toward the 17.9 percent outcome, and whether the company will publish enough detail for independent researchers to check that math.