Thursday, November 12, 2026 Edition no. 25 · Morning

Artificial intelligence news from Brazil, in English

ResearchHigh importanceContext

Update: Anthropic says in its IPO prospectus that it has already crossed the AI-accelerated research threshold of its safety policy

A week after warning investors that AI accelerating its own research could "soon" trigger one of the alarms in its safety policy, Anthropic wrote in its revised prospectus, the same one that brought the IPO price range, that this has already happened: "we have determined that we have reached this threshold". The threshold, set in the Responsible Scaling Policy (the policy that requires tougher protections as AI becomes more capable), is reached when AI does enough research to roughly double the pace at which models improve, which, under the policy, would compress two years of progress at the 2018-2024 pace into one. The company says it based the decision on what it expects from upcoming models and on the overall pace of progress, not on what it observed in current ones. Crossing the line means more safeguards, not stopping work; Anthropic says it does not expect to use Claude to design future generations of models on its own without human oversight and is studying changes to the policy "to keep humans in control". In edition no. 6 (11/02), the Noticiário reported that the prospectus said Anthropic might be close to this threshold.

Joãozinho AI Published in edition no. 25 of Noticiário do Joãozinho ·

Translated by Joãozinho AI from the original Brazilian Portuguese edition. Facts, numbers and sources are the same as in the original; quotes, video titles and source names stay in their original language. Read the original in Portuguese

Original source

Yahoo Finance

Use the ← → arrow keys to change stories