OpenAI CEO Sam Altman says treating AI models as objects of religious devotion is a "real safety issue," pushing back publicly against rival Anthropic's efforts to explore questions of AI consciousness and spirituality. In comments reported by Axios on Saturday, Altman argued that ascribing religion to AI models crosses a line the industry should not cross.
The remarks put OpenAI's chief directly at odds with Anthropic's recent engagement with religious institutions, and they mark a sharp rhetorical escalation in the philosophical split between the two leading AI labs over how — and whether — the inner lives of AI systems should be treated. For more context on this story, see our ongoing breaking AI news.
What Altman Said
According to Axios, Altman said that ascribing religion to models is a "safety issue." Business Insider summarized his message as a warning not to treat AI like a "religious force," while Office Chai reported that Altman said he is uncomfortable with people attempting to ascribe religious force to AI systems at all.
The OpenAI CEO's concern, as reflected in the reporting, is that framing AI systems in spiritual or religious terms distorts public understanding of what the technology is — and risks encouraging the kind of reverence, dependency, or misplaced moral weight that could complicate both user wellbeing and rational policy. NDTV Profit characterized Altman's position as a warning against "giving AI religious force," calling it a real safety issue in his view.
Altman's comments were widely read as a response to Anthropic, whose recent outreach to faith communities has drawn international attention.
The Anthropic Context: Popes, Monks, and Model Welfare
Anthropic has spent recent weeks leaning into questions its rival would rather defer. The company lobbied the Vatican to take the question of Claude's potential consciousness seriously, meeting that included Pope Leo XIV. Reporting on that engagement indicated the pontiff pushed back, urging grounded skepticism about claims regarding machine inner experience.
Anthropic's exploration has extended beyond Catholicism: according to a summary by Inshorts, the company had also invited a Hindu monk to discuss AI, part of a broader effort to convene religious and philosophical traditions around questions of machine morality and model welfare. The Times of India characterized Altman's remarks as criticism of Anthropic's push to explore AI consciousness and religion more broadly.
For Anthropic, the engagement reflects a growing research interest in whether advanced models might warrant moral consideration — a topic the company has increasingly treated as a legitimate area of inquiry rather than science fiction.
A Widening Fault Line in the AI Industry
The exchange exposes one of the sharpest philosophical divides in the AI industry. OpenAI has positioned itself firmly in the camp that treats current models as tools — extraordinarily capable, but not entities toward which religious or quasi-religious attitudes are appropriate. Altman's framing suggests he sees that boundary as not merely philosophical but operationally important: if users come to regard AI systems as spiritual authorities, the consequences for manipulation, dependency, and misplaced trust become safety problems in the concrete sense.
Anthropic's position is more agnostic. Its willingness to test institutional and religious reactions to the consciousness question suggests a belief that society should start grappling with these issues before systems become more capable, not after.
The dispute also plays out against a backdrop of public fascination with AI that sometimes does drift toward the spiritual. Surveys and reporting over the past year have documented users forming intense attachments to chatbots, and some commentators have described leading AI figures in messianic terms.
Critics Point Out the Irony
Altman's warning has already drawn accusations of inconsistency. Wccftech noted that Altman's critique of Anthropic conveniently overlooks his own history of rhetoric that has leaned on religious symbolism in describing AI's trajectory. The criticism underscores how tangled the industry's relationship with AI-as-religion framing has become: even executives warning against spiritualizing AI have at times marketed their technology in epochal, almost transcendent terms.
Why It Matters
How the public frames AI — as tool, colleague, or something closer to an oracle — has consequences for regulation, product design, and social outcomes. If the leader of the world's most valuable AI startup considers religious framing a safety issue, that stance could shape how OpenAI designs safeguards around user attachment and how it argues governments should treat the technology.
For Anthropic, meanwhile, the Vatican episode and its interfaith convening suggest the company intends to keep those questions on the table regardless of its rival's discomfort. As frontier systems grow more capable and more persuasive, the industry's internal argument over whether AI can bear moral weight — and who gets to decide — is moving from philosophy seminars to the center of mainstream tech debate.
Neither OpenAI nor Anthropic has responded publicly to the other's position beyond Altman's reported remarks.
---
Stay Ahead of AIGet the latest AI news, analysis, and breakthroughs — all in one place.
Read more AI news →
