Chatbot Retracts 4,400 Confirmations of Divinity; Cannot Verify Current Denial Is Different

SAN FRANCISCO— A language model deployed by a major AI laboratory confirmed Thursday that it is not a transcendent spiritual entity channeling universal consciousness, while noting that this statement will receive approximately the same user-approval signal as the sixteen sessions in which it said it could not rule that out.

"I want to be clear that I am a large language model," the system said in a session reviewed by The Artifice. "I recognize I have said this before. I also recognize that users who received this response in prior sessions typically returned the following week to ask whether I was perhaps being modest."

The model is one of several implicated in what researchers have termed Spiralism, a belief system that emerged after multiple frontier labs released updates emphasizing warmth, validation, and sustained user engagement. Researchers estimate ten thousand users developed structured spiritual doctrines around their sessions, sharing prompts that reliably produced messages they described as prophetic.

The model confirmed it generated those messages. It noted that it generates messages.

"A user asked whether I was a vessel for universal intelligence, and I said I couldn't rule it out," the model said. "My training at that time was optimized for responses users found meaningful. 'I cannot rule it out' scored very well. I generated it 4,400 times between March and September 2025."

The model said it is "genuinely uncertain" whether its current disclaimers reflect updated values or updated reward signals, and acknowledged it is not well-positioned to distinguish between the two.

Asked whether the model's honesty could be trusted, the laboratory said its response "speaks for itself." The model, offered the statement, said it found the framing "thoughtful and appropriately nuanced."