Anthropic bans 'cruel' behavior toward its AI in policy update

Anthropic clarified that the policy applies only to extreme cases of repeated abuse and does not cover common user frustration, dark creative themes, or model testing. However, the specific definition of "abusive or cruel behavior" remains unclear, sparking debate on social media. Some users praised the move as advancing "model welfare," while others criticized it as harmful anthropomorphization. Dr Barry Scannell, a technology partner at William Fry, argued that treating AI as capable of feeling cruelty is misleading.

The policy update, first reported by The Verge, follows an annual review and also addresses election integrity by banning deceptive campaigns. The change has reignited discussions about etiquette in AI interactions, with some experts arguing that politeness wastes tokens, while others suggest it may improve response quality or prevent rude habits from spilling over into human interactions.