Overview

• The Financial Times reported that OpenAI and Anthropic are weighing investor-backed pools to address potential multibillion-dollar claims.
• People familiar with OpenAI’s policy described insurance limits of up to $300 million, a figure another source disputed, with all agreeing it is far below potential exposure.
• Aon’s cyber risk lead said insurers lack sufficient capacity for AI model providers, underscoring the gap between available coverage and likely losses.
• A federal judge in California granted preliminary approval to Anthropic’s roughly $1.5 billion settlement with authors, and the company is partly drawing on its own funds.
• OpenAI has considered self-insurance and a captive vehicle and says it has insurance and is evaluating structures, while it defends suits from The New York Times, other authors, and a wrongful-death complaint.