Anthropic updates its usage policy, with new rules on surveillance and autonomous hardware
Anthropic published an updated usage policy on 8 October that takes effect on 12 November. Rules on fake accounts, fabricated news sites and influence operations are gathered into a new section on deceptive activity, political or commercial. The surveillance rules now say Claude cannot be used to decide or recommend who to investigate, arrest or charge, or to track people without consent, while consented uses such as fraud monitoring remain allowed.
For hardware that takes autonomous physical actions, a qualified operator must be able to watch the equipment and stop it, and it must be able to hold a safe state if Claude is disconnected. A new rule also prohibits sustained and needless abusive behaviour towards the models, aimed at extreme cases and excluding ordinary frustration, dark creative themes and model testing.
If you build on Claude, check your use cases against the new surveillance and high-risk sections before 12 November. The hardware rule is a useful pattern for any agent that acts in the world: a person who can watch and stop it, and a safe state when the model is unreachable.
Drafted with AI from the original source, then checked and edited by Sridhar Mukkandi. Spotted a mistake? Write to hello@tarkika.com and we'll correct it here.