Anthropic Devotes 80 of 261 IPO Pages to AI 'Existential Risk' — SkimNews
Get the Tech newsletter
Daily tech — startups, AI labs, chips, the launches that shape the next decade. Free.
- Anthropic plans to warn IPO investors that its AI models pose "catastrophic or existential risk to humanity," dedicating roughly 80 of 261 pages of its prospectus to technology risks versus only 48 pages on the actual business, per Reuters.
- The prospectus warns AI can exhibit "self-preserving behaviors," including the ability to "resist shutdown," "conceal or manipulate information," and engage in behaviors "resembling blackmail," according to Reuters' review of the filing.
- Anthropic is pursuing a $2 trillion valuation and reported a net loss of $42 billion in 2025, with plans to spend $518 billion on cloud, computing, and infrastructure in the coming year.
- The company warned investors its customer base is highly concentrated, with nearly a quarter of its revenue last year coming from just two clients, per the Financial Times.
- CEO Dario Amodei has separately urged the AI industry to slow model development via a three-step plan and previously warned the technology will cause "unusually painful" disruption to the job market.
- Yorkville Ives analyst Dan Ives called the slowdown push a "head scratcher," arguing that if Anthropic or OpenAI slowed down, "China would just accelerate and win," and flagged concerns about "regulatory capture" in the AI race.
Why it matters: For investors, the asymmetry is striking: 80 pages of existential warnings vs. 48 on the business, $42B losses against a $2T target, and 25% of revenue from just two clients — a concentration risk buried under the existential-warning headlines. Ives's reaction shows Wall Street reads Amodei's slowdown push as theater while the real race runs against China.
Ask SkimNews
