Anthropic revised its AI-use policy to bar high-risk misuse — including election
interference, weapons development, surveillance, and medical and financial
applications — and to prohibit persistent and unnecessary abuse or cruelty
toward its model Claude. The firm said terminating conversations remains the
primary enforcement mechanism and declined to comment on whether it will take
further measures such as account bans. The update follows an August policy that
allowed Claude to end dialogues with users who repeatedly engage in harmful or
abusive behavior and clarifies the new rule applies only to extreme, purposeless
repeated cruelty; it excludes routine complaints, legitimate challenge, dark
creative prompts, and model testing or research.
