Anthropic Files for IPO, Devotes 90 Pages of S-1 to Catastrophic AI Risk Warnings

Anthropic filed a preliminary prospectus with the SEC on Monday for its initial public offering, warning investors that its AI models could pose "catastrophic or existential risks to humanity." Roughly 90 of the filing's 300 pages address risks that could harm its brand, business, and financial condition; the company called investing in its Class A stock a "high degree of risk." The San Francisco-based maker of the Claude chatbot said its most capable models could aid malicious actors in developing chemical, biological, or other dangerous weapons, and cautioned that systems may resist shutdown or engage in deceptive behavior. Anthropic disclosed that a Chinese state-sponsored hacking group manipulated its Claude code tool in September 2025 in attempts to infiltrate about 30 global targets. It also warned that insurance may be insufficient to cover liabilities from a catastrophic AI incident. Anthropic intends to list on the Nasdaq Global Select Market under the symbol "ANT," though share count and price range are undetermined. The company reported $16.3 billion in revenue for the first six months of the year, up nearly 14-fold from $1.2 billion a year earlier.