Anthropic is introducing significant revisions to its usage policies, marking the first major update in over a year. The changes address emerging high-risk misuse scenarios, including election interference, weapons development, surveillance, and applications in health and financial sectors.
A central component of the new policy prohibits users from engaging in “sustained and needless abusive or cruel behavior” toward Claude. This rule reinforces measures announced last August, when the company revealed it would empower the AI to end conversations with users exhibiting persistent harmful conduct as part of its “model welfare” research initiatives.
According to the updated guidelines, terminating sessions remains the primary enforcement mechanism for such violations. Anthropic did not provide additional commentary on the specific implementation details of this policy shift.
Leave a Reply