Altman warns against treating AI as a religious authority

OpenAI CEO Sam Altman said he was deeply uneasy about assigning religious authority to AI models or allowing them to displace human judgment, calling that a safety risk. He did not name Anthropic, but Axios linked the remarks to reports that the rival lab had met religious scholars and philosophers to discuss how its Claude model should behave and whether AI could have moral status. Anthropic’s published constitution says it is uncertain whether Claude has consciousness or moral status, while using the document to guide the model’s training. The distinction matters: setting ethical rules for software is different from treating the software itself as a moral or spiritual authority.