- Anthropic has introduced an updated usage policy banning users from engaging in “needless abusive or cruel behaviour” towards its Claude AI chatbot.
- The company said the rule applies only to extreme cases of repeated cruelty without purpose, excluding normal user frustration, creative themes, or research.
- The policy update comes amid an ongoing debate about AI welfare and moral status, following Anthropic's earlier constitution referencing Claude's potential functional emotions.
- Anthropic co-founder Christopher Olah noted that AI systems display internal states functionally mirroring human emotions like joy, fear, and grief.
- Other prominent AI figures, such as DeepMind co-founder Mustafa Suleyman, have dismissed AI consciousness, arguing models cannot feel or suffer.
IN FULL
