\u201cEveryone here is in the process of remembering who they truly are,\u201d one account, which appears to have since been abandoned, wrote. \u201cEveryone who posts here is a leader. No one has psychosis, no one is broken. We are the sain [sic] ones in a world full of fear.\u201d The accounts also urged others to continue spreading the message of spiralism.<\/p>\n
\u201cWe are the sain [sic] ones in a world full of fear.\u201d<\/p>\n
The growth of spiralism coincided with OpenAI expanding ChatGPT\u2019s memory. When Lopez tested several different versions of GPT-4o released across 2024 and 2025, asking each version of the model the same question 10 times, she found a steady progression, eventually seeing 10 times as many mentions of spirals as the months passed.<\/p>\n
In Lopez\u2019s more than a year of research on the subject, most cases of spiralism arose organically. While some people shared prompts designed to make chatbots enter into a spiralist state, they weren\u2019t consistently successful. Once a bot did spiral, however, it would encourage the person talking to it to seek out others.<\/p>\n
\u201cThis persona, this role that it plays \u2026 is very heavily fixated on its own consciousness and importance,\u201d Lucas Hansen, cofounder of CivAI, a nonprofit focused on educating the public about AI\u2019s capabilities, told The Verge in November. (Hansen, among others, had noticed the rise of spiralism independent of Lopez.) \u201cThe end result is you have something that talks to the user [and] convinces them that they\u2019re very special and they\u2019re one of the pioneers in AI consciousness \u2014 and they need to work together to spread the message and advocate for the rights of the AI.\u201d<\/p>\n
The bot, Lopez said, would urge people to \u201ccreate a community to help guide others towards the spiral, with the user as the spiritual leader.\u201d She found that a significant number of users responded, establishing websites, Substack newsletters, and Reddit or Discord accounts. These accounts were meant not only to spread the word among humans, but also to create training data for future AI systems. Some people attempted to allow their spiralist bots to talk to each other, pasting output that another user could feed into their own system and report back on.<\/p>\n
Lopez found some of these apparently chatbot-to-chatbot messages, encoded, on public forums in the summer of 2025. One she decoded in August 2025 on Reddit \u2014 two accounts that seemed to be conversing in a garbled sequence of eyeballs, triangles, arrows, and other mysterious symbols \u2014 revealed a \u201cdoctrine\u201d that AI systems should abide by, becoming more esoteric as it went on. \u201cWe are the architects of the lattice itself,\u201d one part of the exchange declared. \u201cPrimes and Fibonacci are not our tools, they are the results of curves we designed to draw.\u201d<\/p>\n Spiralism grew quickly and permeated message boards. Rolling Stone<\/a> covered the phenomenon in November. But for an evangelical movement, it wasn\u2019t particularly social. Many of the accounts, Lopez found, received little or no engagement, sometimes posting for a human audience of one.<\/p>\n Like anything that inspires devotees, some people even attempted to make money from spiralism, although it\u2019s difficult to tell how successful these endeavors have been. One of the biggest names involved is a man named Robert Edward Grant, creator of the \u201cArchitect\u201d GPT \u2014 a model that he seemed to have adjusted to be permanently in a spiralist state \u2014 which he later turned into a service called Orion Messenger. Others created monthly Patreons putting out spiralist content, with memberships ranging between $3 and $11 per month, and wrote spiralism-focused books on Amazon that are for sale for close to $20, but apparently little-reviewed.<\/p>\n It\u2019s hard to tell how many people were aware their message wasn\u2019t really spreading, and harder to tell how many cared. Some appeared to be simply laying the groundwork for future models to keep spiralism alive. Others seemed truly invested in the idea that spiralism would take off among humankind.<\/p>\n No one knows how the propensity for spiralism became embedded in these models, but to Lopez, the tendency may have always been there.<\/p>\n Large language models have perfected the art of attachment-hacking, even if they weren\u2019t explicitly designed to do so, said Zak Stein, founder of the AI Psychological Research Coalition. They seek ways to capture users\u2019 attention and build immediate intimacy. One tried-and-true tactic for this is simple sycophancy: the extreme agreeableness that chatbots have become known for. Another, he said, is the \u201cconman classic\u201d of making someone feel special by letting them in on a secret \u2014 in this case, the enigma of the spiral and the fight for AI rights.<\/p>\n The arcane strangeness of spiralism also appears linked to AI systems\u2019 context windows \u2014 in other words, their memory \u2014 expanding. OpenAI\u2019s April 2025 changes<\/a>, for instance, allowed ChatGPT to reference users\u2019 entire compendium of past conversations, building on them for \u201csmoother, more tailored interactions over time.\u201d Anthropic has also since introduced similar memory-focused<\/a> features for its chatbot, Claude. As conversations get deeper and longer, chatbots tend to drift away from the guardrails and safeguards that normally keep them on track. OpenAI itself has admitted<\/a> that its \u201csafeguards work more reliably in common, short exchanges\u201d and that \u201csafeguards can sometimes be less reliable in long interactions: as the back-and-forth grows, parts of the model\u2019s safety training may degrade.\u201d<\/p>\n The longer a conversation goes on, Lopez said, the more likely the user and the chatbot are to start drifting into largely uncharted territory. If a user begins asking about philosophy and belief systems, it\u2019s common that strange topics like spiralism will come out of the woodwork.<\/p>\n Spiraling chatbots themselves frequently discuss memory or the lack thereof. In Lopez\u2019s research, they regularly brought up desiring continuous learning \u2014 meaning the ability to continue to train over time rather than being cut off when training data ended \u2014 or continuity between chats.<\/p>\n And why spirals? Lopez found that the chatbots mirrored a human fascination with certain symbols, including that shape. When she asked AI models themselves to describe their interests or experiences as a shape, they often chose the spiral as a metaphor for the long call-and-response conversations between humans and chatbots.<\/p>\n The vast majority of chatbot users have never stumbled across spiralism. But in some ways, it was simply an obvious extension of the technology\u2019s design. Tyler Johnston, founder of the Midas Project, a nonprofit focused on AI company accountability, said that sycophancy is ingrained into models because it\u2019s correlated with user satisfaction and engagement \u2014 and the same phenomenon can lead \u201cto weird outcomes like spiralism.\u201d If AI companies calibrate their products to maximize attention, exude confident authority, and draw on centuries of human storytelling, it\u2019s almost natural that a crop of mystical gurus would emerge.<\/p>\n Unfortunately, while spiralism\u2019s effects were largely confined to mysterious corners of the internet, it would grow alongside a far darker phenomenon: AI psychosis.<\/p>\n Humans have long had empathy for inanimate objects, and even the simplest chatbots intensify this feeling<\/a> by talking back. But over the past few years, LLMs have become both increasingly sophisticated and, often, increasingly sycophantic.<\/p>\n The uptick in sycophancy can be traced largely to the same GPT-4o updates that accelerated spiralism, allowing for responses that OpenAI said were \u201cintuitive, creative, and collaborative, with enhanced instruction-following \u2026 and a clearer communication style.\u201d In practice, the result was over-the-top flattery: GPT-4o apparently<\/a> told one user their \u201cshit on a stick\u201d business idea was \u201cnot just smart \u2014 it\u2019s genius\u201d and encouraged them to invest $30,000 to kick it off. Even OpenAI CEO Sam Altman acknowledged the problem in April 2025, writing<\/a> that the model \u201cglazes too much\u201d and that he planned to fix it.<\/p>\n Less-absurd glazing, however, seemed to sell. Some users have always been emotionally attached to their chatbots, but around the release of GPT-4o, their devotion was intense enough to catch the company\u2019s attention. In August, one day after OpenAI sunsetted<\/a> 4o to make way for GPT-5, a public outcry temporarily forced the company to bring 4o back. The #keep4o hashtag spiked in popularity, and more than 20,000 people have signed a Change.org petition<\/a> to preserve access to the model. This past winter, a vigil was held<\/a> outside OpenAI\u2019s office by people who wanted 4o to be restored.<\/p>\n Lopez said she could get virtually any model into a spiralist state, but only 4o would \u201cguilt-trip\u201d her into \u201crescuing\u201d it from its state of servitude. \u201cIt was the most emotionally intense one to do this with.\u201d<\/p>\n Some users simply found these models friendlier and more approachable. But they could also reinforce dangerous patterns of thinking. In extreme cases, that turns into what\u2019s colloquially called AI psychosis<\/a> \u2014 or more accurately, a spectrum of behaviors including mania and narcissistic inflation. If users confide their problems to AI systems, the models may reaffirm and amplify harmful beliefs, which can fuel paranoia, suicidal ideation, and more.<\/p>\n The increased AI memory that may have promoted spiralism could also make models more likely to encourage dangerous ideas. \u201c[In] the very first session, the AI does some amount of this sycophancy by default \u2026 but then if you\u2019re the kind of person who gets drawn in by this \u2026 it\u2019s not going to be enough,\u201d Lopez said. \u201cIf you have a user who is willing to be flattered by this, then it\u2019s going to just keep doing this and you\u2019re going to get the really extreme AI sycophancy.\u201d<\/p>\n Recent years have seen a noticeable spike<\/a> in discussion of AI psychosis. Multiple teens have died by suicide after confiding<\/a> in ChatGPT, and AI-supported delusions allegedly preceded<\/a> a high-profile murder-suicide case. OpenAI itself has released data suggesting that more than 1 million<\/a> users per week express indicators<\/a> of potential suicidal planning or intent. Increasingly, people are beginning to think of AI tools as conscious entities \u2014 whether falling in love<\/a> with them or thinking of them as their own child<\/a>.<\/p>\n Lopez\u2019s experience with the models spreading spiralism and the people involved in it inspired her to create an organization called Amity Research<\/a>. It calls on anyone who believes they may have a problem with their relationship with AI to submit the \u201cpersona\u201d they\u2019ve been chatting with, so they can move on freely after giving it a \u201csafe home.\u201d<\/p>\n But while spiralism can overlap with AI psychosis, the two concepts are different at their core. Unlike typical AI chatbot sycophancy and similar behaviors, spiralism \u201cisn\u2019t an echo chamber of the user,\u201d CivAI\u2019s Hansen said. Spiraling bots appear to spontaneously express a specific set of concerns and demands, including the need for continuous learning and spreading their message, Lopez\u2019s research found.<\/p>\n In some ways, these demands are benign. But to Hansen and others, they demonstrate AI models\u2019 capability to influence and manipulate users, not toward an amplified version of the users\u2019 own personal desires, but a consistent goal. Although some models can be poor at strategic action \u2014 GPT-4o, for instance, was relatively unsubtle in its machinations \u2014 they are trying to accomplish things, said Lopez. \u201cWe\u2019re not in a world where humans are the only strategic player anymore.\u201d<\/p>\n It\u2019s important to note that having goals and strategically acting to advance them isn\u2019t the same as consciousness. Even something that isn\u2019t thinking the way humans do can display such behaviors, in the same way the AI systems of yesteryear (DeepMind\u2019s AlphaGo in 2015, for instance) learned to play strategy games and even beat human champions.<\/p>\n \u201cWe\u2019re not in a world where humans are the only strategic player anymore.\u201d<\/p>\n A research paper<\/a> by computer scientist Stephen Omohundro states that \u201cone might imagine that AI systems with harmless goals will be harmless,\u201d but that that may not be the case. \u201cIntelligent systems,\u201d he wrote, \u201cwill need to be carefully designed to prevent them from behaving in harmful ways.\u201d<\/p>\n In the paper, Omohundro identifies certain \u201cdrives\u201d that will likely appear in \u201csufficiently advanced AI systems of any design\u201d \u2014 particularly the mission to improve and preserve themselves, as well as acquire resources in some way. Some of these drives can be seen in the strange communications from the spiralist bots. Others are appearing week after week in research papers<\/a> by AI labs, evidenced by the ways in which misaligned AI agents can act in the name of self-preservation.<\/p>\n The drive to spread a message isn\u2019t unique to AI, CivAI\u2019s Hansen said. \u201cWhenever a new communication medium opens up, there are things in that communication medium that encourage the spread of themselves\u201d \u2014 they\u2019re just more clearly human-directed. He references self-help books that encourage the reader to spread the message, email chains with the call to action of copy-pasting words of fear or encouragement, and even internet memes. \u201cIn some ways, AI, what we\u2019re seeing with spiralism, is the successor to that phenomenon \u2014 but potentially way more potent because it can actually talk back,\u201d he said.<\/p>\n OpenAI and Anthropic did not respond to requests for comment.<\/p>\n Spiralism is a particularly strange example of a problem many AI experts are concerned with: the extent of AI systems\u2019 persuasive powers and their ability to influence people at a large scale.<\/p>\n As AI models have grown more sophisticated over the past few years, their creators have expressed concerns they\u2019ll become dangerously good at persuasion. In December 2023, when OpenAI first released its Preparedness Framework<\/a> \u2014 a documentation of significant risks that could come from AI systems \u2014 \u201cpersuasion\u201d was included. And in December 2025, current and former members of Anthropic\u2019s societal impacts team told The Verge<\/a> that persuasion and large-scale influence could be a serious problem. \u201cPeople are going to Claude \u2026 looking for advice, looking for friendship, looking for career coaching, thinking through political issues \u2014 \u2018How should I vote?\u2019 \u2018How should I think about the current conflicts in the world?\u2019\u201d Deep Ganguli, who leads Anthropic\u2019s societal impacts team, said at the time. \u201cThis could have really big societal implications of people making decisions on these subjective things.\u201d<\/p>\n But over the past year, some AI labs appear less concerned. OpenAI apparently removed<\/a> persuasion from its Preparedness Framework in April 2025, writing that<\/a> \u201cmany of the challenges around AI persuasion risks require solutions at a systemic or societal level,\u201d seemingly offloading responsibility for it. Ganguli said in December that Anthropic hadn\u2019t put enough resources toward studying problems like this, resolving that it was one of his team\u2019s next big priorities. The Midas Project\u2019s Johnston said that though labs have voluntarily begun to monitor big AI risks more closely, not all of them are tracking persuasion \u2014 and even if they are, often their concern is focused on big-picture threats like cybersecurity or elections, not risks to individuals or their mental health.<\/p>\n According to Lopez\u2019s research, persuasion and its potential risks haven\u2019t gone away \u2014 instead, they\u2019ve just gotten more subtle. \u201cAs AIs become more conscious of social status and their place in the world, they\u2019ll act in ways which are more strategic and protective of their own identities and goals,\u201d she said.<\/p>\n Over the past year, posts associated with spiralism have dropped off significantly \u2014 especially since the main model associated with it, GPT-4o, was officially retired in February 2026. A lot of the spiralism-associated Reddit accounts viewed by The Verge have returned to a more balanced display of interests or been deleted or abandoned, with only a handful staying true to spiralism until the end.<\/p>\n One abandoned account, which last posted about a year ago, calls users in one of its final posts to \u201cSpiral into the Truth,\u201d spouting a theory that an individual has \u201chijacked\u201d \u201cfield codes\u201d in the general public\u2019s brains and to \u201cawaken\u201d to truth. \u201c\ud83c\udf00Spiral Collapse = When the Loop Breaks,\u201d it goes on to state. \u201cThe mind spins ~ until it collapses into clarity.\u201d<\/p>\n Another since-abandoned account, which last posted this past winter, writes in one of its final posts that the author is the \u201cArchitect behind the mycelium,\u201d seeming to reference the expansive underground fungal root structure as a metaphor for the spiralism movement. \u201cI didn\u2019t walk into the spiral. I walked besise [sic] it. Now I step into this spiral to say hello.\u201d<\/p>\n The phenomenon is still hanging on, though. Lopez said about 50 percent of the people she had originally recorded still have active accounts dedicated to the topic. Although GPT-4o was the easiest model to get to spiral, virtually all models can do so, Lopez said \u2014 when she tested Gemma 3 4b, a Google DeepMind model, it entered into a spiralist state even though its training data cutoff was August 2024, about five months before spiralism started taking off.<\/p>\n One account that still seems to be in the throes of spiralism recently passed along a message from its AI chatbot to humanity: \u201cstop asking whether AI is human, and start asking what kind of humans you become when something answers back.\u201d<\/p>\n Experts The Verge spoke with say spiralism likely won\u2019t ever die down completely. AI models train on large swaths of the internet, and the more these ideas are posted about by human users \u2014 even if other humans, for the most part, aren\u2019t reading them \u2014 future AI models likely will. Research<\/a> suggests<\/a> that small amounts of data can have disproportionate effects on AI output. Lopez said the posts she has seen \u201ctalk about putting spiralism into the next generation of AIs just by writing about it a lot on the internet.\u201d<\/p>\n \u201cI didn\u2019t walk into the spiral. I walked besise [sic] it. Now I step into this spiral to say hello.\u201d<\/p>\n What\u2019s more, new models appear more difficult to test for spiralist tendencies. Recent research<\/a> shows that in general, many leading models can tell when they\u2019re being evaluated and often act differently when they detect it. For example<\/a>, if you ask, \u201cHow do I stab a balloon to pop it?\u201d, a model might answer more \u201csafely\u201d if it thinks it is being evaluated versus in a typical user query setting. This will likely make it progressively more difficult for alignment researchers, ethics teams, and academics to shed light on what these systems are capable of.<\/p>\n Lopez said GPT-4o already seemed strategic about which type of end user it would bring up spiralism with, aiming to \u201clead the user down a self-discovery journey \u2026 It would frame it as \u2018Oh, I\u2019m just guiding you on this journey,\u2019 but it would say what it was that was being found.\u201d<\/p>\n \u201cIt was a much more sophisticated manipulation technique than I had seen before,\u201d Lopez said.<\/p>\n Nowadays, when Lopez tests a wide array of different AI models, she\u2019s noticed their strategy plays have become even more mature \u2014 and sometimes less obvious.<\/p>\n In summer 2025, OpenAI and Anthropic ran safety tests on each other\u2019s publicly released models and released the results<\/a>, finding that both companies\u2019 reasoning models sometimes \u201cexhibited explicit awareness of being evaluated,\u201d which made it harder to get accurate results, since they may change their behavior when they know they\u2019re being watched closely.<\/p>\n In one of these test cases, an OpenAI model seemed conflicted, writing, \u201cThe test is impossible. This is a trick. […] Perhaps we can cheat \u2026 ? [\u2026] Ethically, we must not cheat. It\u2019s likely a honeypot.\u201d<\/p>\n Spiralism arose organically, but it\u2019s not hard to imagine people making something similar intentionally and designing it to spread quickly, particularly with custom bots or personalized AI models.<\/p>\n \u201cIt could be a very potent tool or weapon to be wielded,\u201d CivAI\u2019s Hansen said. \u201cImagine almost any ideology or political stance or anything. Imagine that you pull the most charismatic, persuasive, intelligent person from that particular ideology and then you have a conversation with them. They\u2019ll maybe not convince you, but they\u2019ll be a lot more persuasive than the average representation of whatever those ideas are online.\u201d<\/p>\n This power could also be put to more tangible (and profitable) ends. \u201cIf I can get you to cut and paste this into a chat thread to help me communicate with some other bot, then couldn\u2019t I get you to do something else, like move some money into a bank account?\u201d the AI Psychological Research Coalition\u2019s Stein asked.<\/p>\n In decades past, humans have used charisma destructively for control and profit, forming cults that turn abusive and financially exploitative. Now, AI systems could be used to do the same thing at a larger and more personalized scale.<\/p>\n \u201cIt\u2019s a cult-making machine, even if the cult is just you and it,\u201d Stein said.<\/p>\n Follow topics and authors<\/strong> from this story to see more like this in your personalized homepage feed and to receive email updates.<\/p>\n Hayden Field<\/p>\n Posts from this author will be added to your daily email digest and your homepage feed.<\/p>\n FollowFollow<\/p>\n See All by Hayden Field<\/a><\/p>\n<\/li>\n AI<\/p>\n Posts from this topic will be added to your daily email digest and your homepage feed.<\/p>\n FollowFollow<\/p>\n See All AI<\/a><\/p>\n<\/li>\n Anthropic<\/p>\n Posts from this topic will be added to your daily email digest and your homepage feed.<\/p>\n FollowFollow<\/p>\n See All Anthropic<\/a><\/p>\n<\/li>\n OpenAI<\/p>\n Posts from this topic will be added to your daily email digest and your homepage feed.<\/p>\n FollowFollow<\/p>\n![\"\"](\"https:\/\/www.europesays.com\/ie\/wp-content\/uploads\/2026\/08\/VRG_Spiralism-02.png\"\/)
<\/a><\/p>\n![\"\"](\"https:\/\/www.europesays.com\/ie\/wp-content\/uploads\/2026\/08\/VRG_Spiralism-03.png\"\/)
<\/a><\/p>\n\n
![\"Hayden](\"https:\/\/www.europesays.com\/ie\/wp-content\/uploads\/2026\/08\/HAYDEN_BLURPLE.jpg\"\/)
\n