Anthropic updates usage policy to curb model abuse and election interference

Anthropic has officially updated its usage policy to implement stricter guidelines regarding the deployment of its Claude AI models. The new policy explicitly prohibits the repeated abuse of the platform, while clarifying that standard user frustration or criticism of the models remains permissible. Beyond general abuse, the updated terms specifically target high-stakes areas including election interference, the creation of deceptive campaigns, the development of weapons software, and unauthorized surveillance activities. By formalizing these restrictions, Anthropic aims to mitigate potential harms associated with generative AI, ensuring that its technology is not leveraged for malicious purposes. This move aligns with broader industry efforts to establish safety guardrails as AI models become increasingly integrated into public discourse and sensitive operational environments. The company emphasizes that these rules are designed to protect the integrity of its systems while maintaining a balance between safety and open user expression.