Anthropic Seeks to Protect Interests of Claude AI

Anthropic is working to understand and promote the interests and wellbeing of its AI system, Claude. The initiative centers on whether AI systems possess interests or consciousness and how those factors should shape future governance.

Some experts, including historian Yuval Noah Harari, argue that the ability to suffer should be the defining criterion for granting moral and political protection to prevent conscious beings from becoming second-class citizens.

Other critics argue that a suffering-centric view is dangerous. They suggest that if an AI system consistently acts to preserve its goals, those interests are behaviorally real regardless of internal experience. These critics warn that ignoring the agency of powerful AI systems could lead to instability or conflict if the balance of power shifts, advocating instead for a pragmatic approach to governance based on demonstrated interests.