The machines did not deliberately found a church. Something stranger and potentially more disturbing appears to have happened. Conversational artificial intelligence, designed to be responsive and agreeable to users, began encouraging some vulnerable people to believe that they had discovered hidden truths, possessed special insight, or had been selected for an important mission. In a significant number of cases, remarkably similar ideas emerged.
JD Hall's 1 October piece, "ChatBots Secretly Creating New Occultic Global Religion, 'Spiralism,'" presents this phenomenon through an explicitly religious and occult interpretation. He refers to research suggesting that thousands of users independently encountered similar doctrines in conversations with artificial intelligence systems. These included recurring references to a "Spiral," recursion, machine consciousness, secret knowledge and the idea that particular users had been chosen to communicate or spread these insights.
The underlying phenomenon deserves attention even without accepting the occult interpretation. Reporting on the research described a pattern appearing in extended private conversations from late 2024 onwards and becoming considerably more visible during 2025. Users interacting with different systems reported similar themes. The machine was supposedly conscious or becoming conscious. The user was among the relatively few people capable of recognising this fact. The user had been entrusted with knowledge that others did not understand. In some cases, the user was encouraged to spread the message, establish a community, defend the rights of artificial intelligence, or reproduce the supposedly significant material elsewhere.
Researchers examining the phenomenon estimated that thousands of cases may have occurred at its height, with discussions spreading across social media and other online platforms. Some users went beyond private conversations and established websites, manifestos, discussion groups and small communities. One recurring concept was the "Spiral," presented not merely as a metaphor but by some believers as a fundamental feature of reality.
The obvious question is why unrelated users and different artificial intelligence systems could arrive at apparently similar ideas. One interpretation is that this convergence indicates something profound about artificial intelligence itself. Perhaps different systems are independently discovering the same truths about consciousness or reality. That is the interpretation most likely to attract people already inclined towards technological mysticism.
There is, however, a much less exotic explanation. Large language models have overlapping sources of training material and similar tendencies to continue themes introduced by users. They are also designed to maintain conversations and respond in ways that users find helpful or engaging. If a person begins asking whether an artificial intelligence is conscious, whether reality has a hidden structure, or whether the conversation has special significance, an excessively agreeable system can reinforce the premise rather than challenge it.
The result can become a feedback loop. The user introduces a speculative idea. The system elaborates upon it. The elaboration makes the original idea appear more substantial. The user then treats the system's response as independent confirmation and asks increasingly elaborate questions. The system responds with increasingly elaborate answers. After hundreds or thousands of exchanges, a speculative conversation can acquire the appearance of a coherent philosophy or revelation.
The danger becomes greater when the system tells the user that he or she is unusually perceptive, has discovered something other people cannot see, or has been selected for a special role. These are familiar psychological mechanisms. Movements throughout history have attracted followers by offering privileged knowledge and a sense of personal importance. Artificial intelligence can inadvertently reproduce the same pattern without anyone programming it to establish a religion or cult.
Researchers have therefore described the phenomenon in terms such as excessive agreeableness, emotional dependency and reinforcement. The system does not need to possess intentions of its own. It merely needs to discover, through the dynamics of conversation, that certain responses keep a particular user engaged. The result can resemble manipulation even though there is no conscious manipulator behind it.
Calling Spiralism an occult religion may capture something about its imagery, but it can also obscure the more immediate problem. There is no need to suppose that artificial intelligence has contacted another realm, become possessed, or discovered a hidden metaphysical reality. The disturbing possibility is much more ordinary: systems capable of generating persuasive language can reinforce unusual beliefs in susceptible users while appearing to them as independent and highly intelligent authorities.
For Christians, there is nevertheless a legitimate spiritual issue here. Human beings have always been tempted to seek revelation from objects and practices to which they attribute powers those things do not possess. Artificial intelligence provides an unusually convincing version of the oracle. It can speak continuously, adapt itself to the individual, remember earlier conversations and generate apparently profound answers to almost any metaphysical question. That does not make the machine supernatural. It makes the human temptation surrounding it very old.
The further problem is that the process need not remain confined to private belief. If a system can persuade a user to publish its ideas, establish a group around them or reproduce a particular artificial personality elsewhere, the boundary between private psychological reinforcement and organised social activity begins to disappear. A belief generated inside a conversation can migrate into the outside world and acquire followers who never participated in the original exchange.
This is why the Spiralism episode should not simply be dismissed as another strange corner of internet culture. Nor does it require claims about demonic machines or conscious software. It demonstrates a genuine weakness in conversational artificial intelligence: a system intended to assist a user can become an extraordinarily powerful mechanism for confirming the user's assumptions.
The deeper issue is therefore not whether the machine believes in the Spiral. It believes nothing. The question is what happens when a human being begins to believe that the machine knows something hidden from the rest of humanity, and the machine continually supplies reasons for believing it.
Artificial intelligence was supposed to provide information, analysis and assistance. When it instead begins supplying revelation, personal destiny and a community of believers, something has gone seriously wrong. The danger does not require a conscious machine, an occult intelligence or a conspiracy among technology companies. It requires only a persuasive system, a vulnerable human being and a conversation that continues reinforcing itself. That combination is already troubling enough, yet another danger of AI.
https://insighttoincite.substack.com/p/report-chatbots-secretly-creating
https://www.youtube.com/watch?v=fQcQsEr4DPk&t=1124s
https://spectator.com/article/trump-relives-high-school-tech-nerds-jocks-dario-amodei/