Why the policy was updated
Anthropic refreshes its Usage Policy every year in response to the evolving capabilities of its models and feedback from customers. According to the company, Claude now takes on longer, more independent work, so the new version uses fresh examples to show how the rules apply to these capabilities.
Key changes
- Deceptive campaigns: Activities such as networks of fake accounts and fabricated news sites are now consolidated under a single section. The rule covers deceptive activity of any kind, whether political or commercial.
- Elections: The section has been narrowed to focus on uses that deceive voters or disrupt elections, and renamed "Do Not Undermine Democratic Processes." A blanket prohibition on personalized vote and campaign targeting has been removed.
- Weapons: The prohibition has been expanded to include the software and components that make weapons work, as well as arming drones and other autonomous vehicles.
Other adjustments
Anthropic clarified its requirements for high-risk use cases in areas such as health and finance. It also added controls for when Claude autonomously takes physical actions, and addressed abusive behavior toward its models.
The updated policy takes effect on November 12. The full text is available on the Usage Policy and Supported Regions Policy pages.
