Anthropic bans ‘abusive or cruel behavior’ toward Claude
TL;DR
Anthropic is making changes to its usage policy for the first time in over a year to reflect new and high-risk cases of misuse - including election interference, weapons development, surveillance, and health and financial uses. But one of the most significant changes prohibits "sustained and needless abusive or cruel behavior" toward Claude. Last August, the company announced it would allow Claude to end conversations with "persistently harmful or abusive" users as part of its research into "model welfare.
Nauti's Take
For teams building with Claude, the new clause is primarily a governance issue: test suites, red-team prompts, and support automations need a clear boundary between rigorous probing and repeated personal abuse. Before moving a workflow into production, document when a conversation may be terminated, what fallback users receive, and whether Anthropic publishes criteria consistently enough for safety and quality tests to remain reproducible.
Sources