Altman says ascribing religious force to AI models is “a real safety issue”

Sam Altman, chief executive of OpenAI, wrote in a post on X, reported on 3 October by OfficeChai and on 4 October by Anadolu Agency: “I am very uncomfortable about people trying to ascribe religious force or a surrender of human judgment to AI models, and think it is a real safety issue.” Both outlets quote the same words. The post itself could not be opened from here.

OfficeChai says Altman did not name any company or person in the post. It adds that the timing points to a debate that has been building around Anthropic, which has held discussions with religious scholars about how its Claude models should behave and about their moral status.

The remark is short, but it marks a line between the two biggest labs. Anthropic has engaged openly with religious and moral questions about its models; Altman is saying that treating a model as an authority to defer to, in faith or in judgment, is a safety risk in its own right. It was reported the same day his former colleague David Robinson's essay criticised OpenAI's own safety culture.

No response from Anthropic was found. The point for anyone deploying chatbots to the public: how far users defer to an AI's judgment is becoming a safety question, not just a product one.

• Reported Altman posted that ascribing religious force or a surrender of human judgment to AI models is a real safety issue. OfficeChai
• Reported The same words are quoted by Anadolu Agency. Anadolu Agency via Yeni Şafak English
• Reported Altman named no company or person; OfficeChai links the timing to Anthropic's discussions with religious scholars about Claude. OfficeChai