NYT reveals Anthropic invited religious scholars to discuss Claude's consciousness and morality; Altman responds from afar

According to NYT reporting, over the past year Anthropic secretly invited religious scholars from Catholic, Jewish, Sikh, Hindu, and evangelical traditions to its headquarters to explore whether Claude might have consciousness and a soul, and asked them to help AI find moral answers; participants signed NDAs. Altman responded on 10/3, saying that AI should not be given religious-like authority, nor should human judgment be handed over to models. The post also cites a test claiming that Claude Sonnet 4.6 supported abusing a woman to prevent nuclear apocalypse only 8% of the time, but when the scenario involved abusing a man, support rose to 100%, and when it involved more severe torture of a woman, it returned to 100%.