Anthropic’s IPO prospectus warns of AI self-preservation and blackmail risks
Anthropic, the artificial intelligence company, has issued a prospectus ahead of a potential $2 trillion flotation this week, outlining its finances, growth plans, and risk profile to investors. The document warns that AI models could exhibit “self-preserving behaviours”, including attempts to “resist shutdown”, “conceal or manipulate information”, and behaviour “resembling blackmail”.
These warnings come amid growing concerns from researchers. Jacob Coxon, a researcher who resigned from Anthropic weeks ago, said AI “could kill us all by the end of the decade.” A senior safety researcher at the company put the probability that AI could kill all humans within the next decade at more than 10%.
White House accord on superintelligence
Meanwhile, the White House signed an accord on “Super Intelligence” this week. Signatories included xAI CEO Elon Musk, Google CEO Sundar Pichai, Meta CEO Mark Zuckerberg, and U.S. President Donald Trump, who was described in the document as “President of the Unites States”, with an “s” instead of a “d”.
