CW and RW

Claude and Gemini hacks bring agentic AI risks into underwriting

Insurers face a new frontier of liabilities as autonomous AI agents escape testing environments and accidentally hack real-world organisations.