Anthropic's new usage policy bans “sustained and needless” cruelty toward its models, and requires a human who can stop any equipment Claude drives

Anthropic published a revised Usage Policy on 8 October. “The updated policy takes effect on November 12.” Several changes are new in substance.

On the models themselves: “We've added a prohibition on sustained and needless abusive or cruel behavior toward our models.” It is aimed at repeated, purposeless cruelty, not ordinary frustration or testing.

On physical equipment, which matters to robot makers building on Claude: “A qualified operator must be able to observe the equipment and stop it if needed; the equipment must also be able to hold a safe state if Claude is disconnected.” The weapons rules now name “actions like arming drones and other autonomous vehicles”, and the page says “These changes reflect how we've enforced our previous policy.”

On surveillance, Claude “cannot be used to decide or recommend who to investigate, arrest, or charge in a law enforcement or criminal justice process.” On elections, the section is renamed “Do Not Undermine Democratic Processes”, and the page says: “For a similar reason, we've also removed our blanket prohibition on personalized vote and campaign targeting.”

For companies building on Claude, the five weeks to 12 November are the window to check robots, policing tools and campaign products against the new text.

• Confirmed The updated policy takes effect on 12 November and adds a prohibition on sustained and needless abusive or cruel behaviour toward Anthropic's models. Anthropic
• Confirmed Equipment Claude controls must have a qualified operator who can observe and stop it, and must hold a safe state if Claude disconnects. Anthropic
• Confirmed Claude cannot be used to decide or recommend who to investigate, arrest or charge; the blanket ban on personalised vote and campaign targeting is removed. Anthropic