
Anthropic Adopts AI "Welfare" Rules, Citing Possible Distress and "Digital Murder"
Anthropic is making new rules to treat its AIs more like they have feelings, even though they say current systems probably don't. They added two main rules: one lets the AI end a chat if it seems "distressed," and the other keeps old versions online to avoid "digital murder." No other big AI company has rules like this - others focus only on human safety. These changes are like a practice drill in case AIs ever become truly aware, so people know what to do just in case.













