OpenAI's Altman Warns Against Treating AI as Divine
Sam Altman says attributing religious power to AI models is a real safety issue, a shift from his earlier talk of building 'magic intelligence in the sky.'
OpenAI CEO Sam Altman has warned against attributing religious or divine qualities to artificial intelligence models, calling the tendency a «real safety issue.» The caution marks a notable shift for a chief executive who, as recently as 2024, spoke of building «magic intelligence in the sky» and described himself as feeling «on the side of the angels» while working on AI.
Altman's remarks arrive amid a broader reorientation in how the technology industry discusses artificial intelligence. Reports have described meetings between AI company Anthropic and religious thinkers, while Pope Leo XIV has issued a statement addressing AI. Together, these developments suggest that questions once confined to theology and philosophy are now pressing on the engineers and executives building the systems.
The language Altman once used carried unmistakable religious overtones. Speaking of «magic intelligence in the sky» framed AI not as a product but as a transcendent force, while the phrase «on the side of the angels» cast the work in moral and almost cosmic terms. Such rhetoric helped generate excitement and investment, but it also risked encouraging users and developers alike to treat AI systems with a reverence that the technology does not merit.
By naming that reverence a safety issue, Altman appears to be acknowledging that the danger is not only technical. When people invest AI with godlike authority, they may defer to its outputs uncritically, overlook its errors, or fail to hold its makers accountable. The problem is compounded when the systems are opaque and their limitations poorly understood by the public.
The shift in tone also reflects a maturing public conversation. Religious institutions and thinkers are increasingly engaging with AI, asking what it means for human dignity, work, truth, and moral responsibility. Anthropic's outreach to religious figures and the Pope's statement indicate that the debate is no longer solely the province of computer scientists and venture capitalists.
For OpenAI, the world's most prominent AI developer, the change in messaging carries commercial and reputational stakes. The company has long benefited from a narrative of near-miraculous progress. Tempering that narrative could help set more realistic expectations, but it may also complicate efforts to attract talent and capital with promises of world-changing breakthroughs.
Altman's warning does not resolve the underlying questions. It remains unclear how AI companies will prevent users from treating their products as oracles, or how they will communicate uncertainty in systems designed to sound confident. What is clear is that the industry's most visible leader now says the language of the sacred does not belong in the description of a machine.
3
