Anthropic Tightens Rules on Cruel Behavior Towards Its AI Assistant

The company’s new policy prohibits gratuitous cruelty towards its conversational AI, sparking debate on AI consciousness.

Anthropic’s AI assistant, Claude, was asked for its opinion on the policy, and it acknowledged the potential criticisms while highlighting the paradox of asking its opinion on a rule meant to protect it.

Claude stated that the policy is ‘defendable, as long as it is well understood’ and that it is based on a precautionary reasoning due to the uncertainty about its moral status.