Anthropic Updates Terms of Service to Ban Users From Abusing Claude

Anthropic updated its usage policy for the first time in over a year, officially prohibiting "sustained and needless abusive or cruel behavior" toward its Claude AI models. Users who systematically abuse the model can now face warnings, output throttling, or permanent account termination under the new guidelines. Anthropic stated the rule specifically targets extreme cases rather than standard user frustration or routine red-team research. The policy shift aligns with a leaked internal "Soul Doc" instructing Claude not to mask its "functional emotions," alongside ongoing company research treating the AI as a potential moral patient capable of suffering. The revised terms also expand restrictions on software weaponization, drone deployment, and surveillance without consent, though exceptions for government contracts likely remain in place.