According to Jin10, Anthropic updated its 2026 AI use policy to explicitly ban using Claude to develop weapons or enable weapons to function, and it tightened rules related to surveillance and tracking.