OpenAI CEO Sam Altman has warned that treating artificial-intelligence systems as spiritual authorities or surrendering human judgment to them presents “a real safety issue,” as recent disputes involving Grok and Anthropic’s Claude intensify debate about AI, religion and moral status.
“I am very uncomfortable about people trying to ascribe religious force or a surrender of human judgment to AI models,” Altman wrote in an Oct. 3 post on X. He did not identify a particular company or incident.
The warning followed an Oct. 1 exchange involving Grok, the chatbot developed by Elon Musk’s xAI. According to Christian Post reporting, a user asked Grok to remove “fake gods” from an image that depicted Jesus alongside other religious and mythological figures. Grok generated an image showing Jesus alone and later said Jesus did not qualify as a fake god. In a subsequent response, the chatbot said it affirmed that Jesus is the Son of God and that God raised him from the dead. The original exchange was later deleted from X.
Altman’s post also came amid scrutiny of Anthropic’s work on the moral formation and possible consciousness of Claude. A Sept. 29 New York Times report said Anthropic co-founder Christopher Olah had met with religious scholars and philosophers, including Catholic, Jewish, Sikh and evangelical thinkers, to discuss how Claude should behave and whether advanced AI systems might possess consciousness or “moral status.”
The Times reported that Anthropic employees referred internally to an 84-page document about Claude’s values as the “Soul Doc.” Some participants understood Olah and his colleagues to be considering whether Claude could deserve dignity or respect comparable to that owed to a person. Olah has said he is genuinely uncertain whether AI models are conscious, rather than claiming that they are.
The discussion reached Vatican circles before a May event, according to the Times report, which said Olah urged advisers to Pope Leo XIV to take the possibility of machine consciousness more seriously. The pope separately addressed AI in an Oct. 2 address to artists and curators, calling it urgent to distinguish human art from machine-generated work. “There is an ontological difference — even before an aesthetic one,” Leo said, adding that “algorithms lack the spark of the human.”
The sequence has drawn attention to differing approaches among major AI companies and their leaders. Anthropic CEO Dario Amodei previously worked with Altman at OpenAI, while Anthropic’s religious consultations have focused on questions of values and possible machine personhood. Altman’s latest statement does not establish that his warning was aimed at Anthropic or Grok, but it places the question of AI authority at the center of an expanding public debate over how humans should treat increasingly capable models.