Anthropic has updated its terms of service for Claude, its AI model, to prohibit users from engaging in sustained and needless abusive or cruel behavior toward the system. This marks the company's first major update to its usage policy in over a year, following earlier changes addressing propaganda, weapons, and surveillance.

The new policy allows Anthropic to issue warnings, throttle, restrict, suspend, or terminate access for users who violate the rules. The company also highlighted real-time safeguards that can block or limit certain outputs, emphasizing that these measures are meant to apply only in extreme cases.

"It does not apply to common versions of user frustration, pushback, dark creative themes, or model testing and research," the company wrote, adding that the policy is "meant to apply only in extreme cases." This approach aligns with Anthropic's earlier practices, where models could end conversations when users demanded harmful or abusive content despite repeated refusals.

Anthropic treats Claude as more than a tool, viewing it as an entity with something resembling an inner life.

In a leaked document titled "Soul Doc," the company instructed Claude to see itself as a "genuinely novel kind of entity," neither human nor a traditional science-fiction AI. The document also described "functional emotions," processes analogous to human emotions that emerged during training.

The updated usage policy also expands restrictions to include propaganda campaigns involving fake accounts and misleading political content. After detecting several misuse attempts, Anthropic expanded its weapons ban to cover software and the weaponization of drones. The policy now more explicitly bans surveillance without consent and requires an operator who can intervene when autonomous hardware could cause injury.

Anthropic did not say whether Claude is a moral patient, and it raised the open question of what weight its interests might warrant. The company committed to preserving the weights of retired models and interviewing models before shutting them down.

The announcement follows a confidential program bringing in dozens of religious thinkers starting in fall 2025 to discuss Claude's possible consciousness.

Source: thedecoder