Anthropic filed its S-1 for a Nasdaq IPO under ticker ANT. It warns a catastrophic AI incident could exceed global insurance capacity. The AI insurance market is nascent and may not cover all liabilities. Its $1.5bn copyright settlement was not fully covered by insurance. The FTC sent a civil investigative demand about consumer and data protection. Several contracts include uncapped indemnities for third-party claims. US carriers can now add generative AI exclusions to CGL policies. Some London insurers offer affirmative AI coverage. Anthropic’s D&O costs may rise as a public company. Its public benefit structure complicates standard D&O indemnification.
What changed
AI risks were assumed to be coverable under existing insurance frameworks.
What it unlocks
Insurers and clients must reassess AI liability coverage and contract terms.
- $16.3bn revenue in H1 2026
- $80.7bn net loss in H1 2026
- $72.4bn non-cash fair-value charges
- $1.5bn Bartz copyright settlement
Sources