Anthropic lists AI risks in its IPO filing
Anthropic, the AI developer, has filed an S-1 registration statement with the US Securities and Exchange Commission ahead of a planned initial public offering. The filing's risk factors section runs to 93 pages and over 70,000 words, and it warns that near-future AI models could pose catastrophic or existential risks to humanity. The company also cites political opposition to data centers in 16 states as a major risk. Fortune reports the company posted $16.29 billion in first-half 2026 revenue and a net loss of $80.7 billion for the six months to June 2026.
Key facts
- The company plans to spend up to $567 billion on compute.
- Anthropic agreed to buy up to $120.5 billion in compute from SpaceX through 2029.
- Anthropic researcher Evan Hubinger put the chance of AI destroying humanity over the next decade above one in 10.
- The prospectus says Anthropic cannot guarantee its AI safety efforts will be adequate, given the technology's complexity.
Summarised with AI from 5 sourcesRead our AI policy
Why it matters
A company warning in a public filing about catastrophic AI risks gives investors and regulators an official record of how the developer assesses those dangers.
What changed
First published from The Economic Times, CNA, CNBC, Fortune and Phys.org