Anthropic Tells IPO Buyers Its AI May Be About to Speed Itself Up

Anthropic just told the people buying its stock that its own AI might be about to hit a line the company drew for itself. In an amended S-1 filed with the SEC on November 2, the AI lab warned it “may imminently meet” the Responsible Scaling Policy definition of “dramatically accelerated progress attributable to AI” — its tripwire for AI that speeds up the building of the next AI. The company plans to list on Nasdaq under the ticker ANT. No model and no date are named.

The change isn't subtle. The original October 26 filing said Anthropic “may not identify a significant acceleration in our rate of development until after it has occurred.” The amendment rewrites that to “have in the past not identified, and may in the future not identify.” The Gadgeteer notes the swap turns a forecast into an admission — the company is now saying it has already missed a speedup once. The new language also includes a pledge to keep humans in control of frontier model development.

Anthropic calls AI-assisted research “closing the loop,” and it sells the practice as a competitive edge. According to the filing, its models now run experiments, generate and process training data, automate evaluations and safety audits, and write and review code for the models that come after them. The internal numbers are loud: the value of Anthropic's own use of its models, at list prices, grew more than 250 times between January and September 2026. By April, that monthly value had passed what the company spends on its people each month, stock pay excluded.

The bar for tripping the alarm is spelled out. Version 3.4 of the Responsible Scaling Policy, in effect since July 8, counts one route as double the rate of AI progress — “as much progress in one year as one would see in two years at baseline.” Anthropic says it has braked before, delaying model releases and limiting early availability, including with Mythos through Project Glasswing.

There's a flip side in the same risk factor. Lean on AI too little, Anthropic warns, and rivals could leave it “falling progressively further behind.” The company says it doesn't expect to safely let its AI “fully autonomously design and build successive generations of frontier models without meaningful human supervision or understanding, if and when that becomes technically possible,” and it's weighing policy updates to keep humans in charge.

This section of the prospectus isn't optional. Item 105 of Regulation S-K requires companies to spell out what makes an investment speculative, and under Section 11 of the Securities Act, buyers can sue over a material misstatement or omission. Investors won't run the show, though: the Long-Term Benefit Trust, a panel of trustees with no financial stake, elects a majority of the board, and a single Class F share held by the seven co-founders' LLC carries 50.1% of votes on most other stockholder decisions. The prospectus opens with a letter signed by all seven. Watch the word “imminently” — it's sitting in a document Anthropic can be sued over.