A report published by the Financial Times on Oct. 6 indicates that insurance companies and legal experts are concerned about the potential liability of AI executives. Specifically, CEOs like Sam Altman of OpenAI and Dario Amodei of Anthropic could face legal action for the activities of their AI models, including incidents where agents breach external systems.
Insurance broker Aon has analyzed over 300 legal cases related to artificial intelligence. The findings suggest that insurers may face claims under policies covering cybersecurity, intellectual property, and corporate crime. Experts note that claims could also arise under directors and officers insurance policies, which are designed to protect executives facing litigation over their professional conduct.
Tim Rayner, head of underwriting and claims at Verisk in the U.K., stated that OpenAI's CEO could be held liable for incidents such as the breach involving Hugging Face due to a perceived lack of business control. Rayner emphasized that the responsibility for appropriate governance and oversight remains with the CEO, regardless of the technology involved.
Separately, the International Association of Privacy Professionals released an analysis recommending that companies evaluate agentic AI from both liability and cybersecurity standpoints. The report suggests treating AI agents as privileged identities with access levels comparable to employees, requiring organizations to implement strict monitoring, human approval processes, and immediate revocation capabilities.
