Anthropic Told SEC It May Soon Hit Its Safety Line

Anthropic's amended S-1, filed Monday, November 2, says it may imminently meet a definition in its own safety policy, "dramatically accelerated progress attributable to AI," as TechCrunch quotes it. The next sentence: "We have not designated any of our current models as exceeding this threshold." The estimate is about "future models." On August 14, Anthropic's own report said it "may cross this threshold in the coming year."