News

Anthropic Updates Usage Policy to Explicitly Ban Model Abuse, Election Interference

Anthropic has released an updated usage policy for its Claude AI assistant, introducing clearer restrictions on several categories of behavior while maintaining space for legitimate user feedback.

The revised policy explicitly prohibits repeatedly abusing Claude in extreme cases. Anthropic clarified that ordinary expressions of frustration, criticism, or dissatisfaction with the model remain allowed—only systematic or severe abuse crosses the line.

Among the newly articulated restrictions are election interference and coordinated deceptive campaigns. The policy now explicitly bans using Claude to influence electoral outcomes or run inauthentic influence operations. Weapons-related software development and surveillance applications are also explicitly restricted under the updated terms.

The move reflects a broader industry trend toward more detailed usage policies as AI companies grapple with diverse ways their models might be misused. Anthropic's update aims to provide clearer guidance for users while establishing firm boundaries around high-risk applications.

The company joins other major AI developers in tightening policy frameworks as deployment scales across consumer and enterprise products.

Sources