TechCrunch reported that Anthropic updated its usage policy for the first time in over a year to prohibit sustained and needless abusive or cruel behavior toward its Claude AI model.
Terminating conversations remains the primary enforcement mechanism for the company when dealing with persistently harmful or abusive users. Anthropic did not comment on whether additional user bans or enforcement measures would follow.
The company stated that the new rule applies only in extreme cases involving repeated cruel actions with no discernible purpose. Common user frustration, pushback, dark creative themes, and research activities are exempt from the restriction.
Campaigns and weapons
Anthropic gathered existing restrictions into a new ban addressing deceptive commercial and political campaigns. The rules prohibit efforts to obscure message authorship, amplify content through fake accounts, spread voter misinformation, and disrupt elections.
The updated policy also expands previous prohibitions on weapons development to include software and components that operate weapons, alongside actions involving the arming of drones and autonomous vehicles.
Surveillance rules
Anthropic clarified its surveillance restrictions following threat intelligence reports indicating increased use of Claude-powered tracking against political dissidents. Real-time and retrospective tracking without consent are prohibited under the updated guidelines.
Claude cannot make investigation, arrest, or charge decisions in criminal justice processes, nor can the model build or improve surveillance tools.
