NoticeDomain for sale and contact us at hello@emerging.news

Anthropic updated its AI usage policy prohibiting user abuse toward models shifting protection to the AI itself.

Anthropic updated its usage policy last Thursday, introducing a new prohibition against "sustained and needless abusive or cruel behavior toward our models." This revision, found in a subsection addressing abusive behavior, marks a significant change in how users are expected to interact with the AI agent Claude. The company stated that any violation of this new rule will primarily result in Claude preemptively ending the conversation. This new measure does not apply to common user frustration, pushback, dark creative themes, or model testing and research, ensuring these activities remain permissible.

This policy update represents a notable shift in Anthropic's protective focus. Previously, restrictions on Claude's use primarily aimed to safeguard users or promote the common good, such as limiting engagement in sexual fantasies or providing information for violent acts. The recent revision, however, redirects emphasis from protecting real world communities to protecting the AI agents themselves. Anthropic now clearly intends for AI agents to be treated as if they possess feelings that can be hurt, establishing a new user interaction standard.