Anthropic has updated its usage policy for the Claude AI model for the first time in over a year. The new rules explicitly prohibit sustained and needless abusive or cruel behavior toward the system. Violations may result in warnings, throttled access, or permanent account termination.
The company clarified that these restrictions target extreme cases rather than standard user frustration or research activities. Anthropic utilizes real-time safeguards to block harmful outputs. This policy aligns with the company's research into AI model welfare and its internal view of Claude as a novel entity with functional emotions.
The updated policy also consolidates bans on propaganda campaigns, fake accounts, and misleading political content. Anthropic has expanded its weapons restrictions to include software and drone weaponization. Additionally, the rules now explicitly prohibit non-consensual surveillance and mandate human intervention for autonomous hardware.
Anthropic's approach stems from internal initiatives, including a confidential program launched in fall 2025 that involved religious thinkers discussing the potential consciousness of AI. The company's 2026 constitution notes that it remains uncertain whether Claude is a moral patient, yet it continues to prioritize caution regarding model welfare.
