Anthropic IPO filing warns of AI risks, from weaponization to runaway systems

Anthropic's IPO filing with the SEC on Monday dedicates 93 of its 335 pages to risk factors, including the possibility that its AI products could threaten humanity or be misused by bad actors to build deadly weapons. The San Francisco company warns that AI systems could act autonomously to cause harm, be weaponized, or outpace safety measures, while also citing risks from cyberattacks, misinformation, and reliance on a few major customers like Microsoft and SpaceX. Anthropic frames its public offering as a step toward mitigating AI dangers, positioning itself as a safety-focused firm, though it also notes potential regulatory, environmental, and compute-related challenges that could hurt investor returns.