The Next Cyber Insurance Claim May Start With Authorized AI

Jay Bavisi of EC-Council highlights emerging cyber insurance risks from AI agents operating with authorized access, which can cause losses despite following approved workflows. Insurers must assess not just AI models but the level of authority granted to these agents, as increased privileges raise potential damage. Establishing authority budgets for AI agents is crucial to balance utility and risk, with early incidents underscoring the need for updated risk frameworks.