Anthropic updated its usage policy to prohibit sustained and needless cruel behavior toward Claude, with the AI model able to end abusive conversations as the primary enforcement mechanism. The policy reorganization also introduces stricter rules on deceptive campaigns, elections, weapons development, law enforcement misuse, physical harm controls, and non-consensual intimate imagery, taking effect November 12.
Anthropic updated its Usage Policy effective November 12, 2026, clarifying rules on deceptive campaigns, weapons development, surveillance, and abusive behavior toward Claude. The policy consolidates restrictions on influence operations and fake accounts into a new section, refocuses election rules on voter deception and disruption, and expands weapons prohibitions to include software and autonomous vehicle control.