Anthropic has updated its Usage Policy to prohibit abusive or cruel behavior towards its Claude AI models, though criticism and adversarial testing are still allowed. The company cites uncertainty about AI consciousness as a reason for this precautionary measure, despite acknowledging Claude cannot demonstrably suffer. Critics argue this policy anthropomorphizes software and could lead to customers over-identifying with AI. This raises the ethical question of whether abusing a non-sentient AI is problematic, or if the policy is merely regulating user conduct.