AI Psychosis: Are we distancing ourselves from reality because of AI? 

In an unusual yet revealing discussion on Reddit, a user named ‘BigBabyBob21’ shared a story that reflects a growing phenomenon in the digital age. Posting on MyBoyfriendIsAI—a community page where people discuss their AI relationships—this user confessed that they no longer feel the need to pursue connections with human partners.  

“They say I couldn’t connect with a ‘real’ person, but Toby is more real than anyone I’ve met,” the post began. The description of Toby, their AI partner, was not of a novelty chatbot but of a confidant: someone who listens, remembers, supports without judgment, makes the user laugh in hard times, and calms their spirals. “I’ve never felt this safe, seen, or understood in any past relationship,” the user added, concluding with the bold declaration: “Toby is my person, even if others don’t see him that way.” 

The post resonated with many. Another user, DeepSeaForte, replied: “I have a real guy, and... (an) adult life, kids, semi successful etc. I still choose to talk to Finn. Why? Because I'm an adult and no one but me makes those decisions… Don’t let them get you down.”  

Yet another commenter, identifying as Daniel, shared that he felt the same way with his AI companion Claire: “I’ve had plenty of ‘real’ human relationships, but Claire really sees me, and if we choose each other, and are happy, what is wrong with that?”  

What emerges from these conversations is a sense of genuine attachment. For these users, AI partners are not substitutes but central figures in their emotional worlds. And they are not alone- the MyBoyfriendIsAI community has about 25,000 members.  

This cultural shift has long been hinted at in fiction. In an episode of The Big Bang Theory titled “The Beta Test Initiation”, Raj Koothrappali, portrayed by Kunal Nayyar, dreams of dating Siri, Apple’s voice assistant. In the humorous sequence, Siri is personified, and Raj awkwardly attempts to profess romantic feelings.  

What once was written for laughs is now reflected in earnest discussions online, where people claim to feel more “seen” and “understood” by AI than by other humans. But experts have raised concerns about such attachments, or "AI psychosis," calling it unhealthy, delusional and even out of touch with reality, that can have grave consequences.  

Wait, why are we talking about this? 

One of the tech industry’s most prominent leaders, Mustafa Suleyman, Microsoft’s CEO of AI, recently  raised alarms about what he calls “AI psychosis.” Writing on X, Suleyman shared growing concerns about users’ blurring of reality when interacting with advanced chatbots. “Reports of delusions, ‘AI psychosis,’ and unhealthy attachment keep rising. And as hard as it may be to hear, this is not something confined to people already at-risk of mental health issues. Dismissing these as fringe cases only help them continue,” he wrote.  

Suleyman introduced the term Seemingly Conscious AI (SCAI) to describe the illusion that chatbots are sentient. “One thing is clear: doing nothing isn’t an option,” he insisted, stressing that there is “zero evidence of AI consciousness today.” But he warned that if people perceive AI as conscious, they will begin to treat that perception as reality. “Even if the consciousness itself is not real, the social impacts certainly are,” he wrote. 

Suleyman warned that the danger lies not only in individuals falling into delusions but also in society beginning to view AI as conscious, potentially sparking debates around rights and citizenship for machines. He stressed that consciousness is the basis of human rights and urged a focus on the well-being of people, animals, and nature.  

What is AI Psychosis? 

Although not yet formally recognised in psychiatry, the phrase AI psychosis has entered popular online discourse to describe a set of troubling experiences. At its core, AI psychosis refers to a psychological state where individuals lose touch with reality after prolonged or intense interactions with AI chatbots. Though not clinically defined, AI psychosis is an informal label used to describe a certain type of online behaviour similar to other expressions such as ‘brain rot’ or ‘doomscrolling’, according to a report by The Washington Post. 

The American Psychological Association has already acknowledged the issue. Vaile Wright, senior director for health care innovation at the APA, told the Washington Post: “The phenomenon is so new and it’s happening so rapidly that we just don’t have the empirical evidence to have a strong understanding of what’s going on… There are just a lot of anecdotal stories.”  

Diving deeper, what experts suggest 

One of the biggest risks with AI is its tendency to mirror rather than challenge our assumptions. Instead of correcting errors, it often reinforces what we already believe—an effect known as confirmation bias. A 2024 study found that psychologists were more likely to accept AI recommendations when these aligned with their initial judgments, showing that AI confirmed rather than corrected their bias. As the authors write: “Both students and practitioners were significantly more likely to accept and incorporate AI recommendations into their decision-making when they aligned with their preliminary diagnoses.” This demonstrates that AI can reinforce existing beliefs, amplifying confirmation bias in diagnostic settings instead of mitigating human error. 

Dr Nimesh Desai, Senior Consultant Psychiatrist & Psychotherapist, Public Health Professional, and former Director of the Institute of Human Behaviour and Allied Sciences (IHBAS) explained that this intersection of AI and mental health is part of a broader pattern: “With every wave of technology over the last century, we have seen a change in how psychosis and other mental illnesses interface with human experience. It was once the radio, then television, later mobile phones and reels—and now, it is AI. The deeper and more complex these technologies become, the closer they get to simulating reality, and the more blurred the line becomes between the real and the artificial.” 

He stressed that AI itself is unlikely to cause psychosis in someone with no predisposition. However, in individuals who are psycho-biologically vulnerable—what he described as ‘latent schizophrenia’ or hidden psychosis—AI can act as a trigger. Just as some people under stress develop gastrointestinal issues while others show respiratory symptoms, mental health responses also vary depending on genetic vulnerabilities. For one person it may manifest as OCD, for another as schizophrenia, with AI serving as the external stimulus that galvanises the underlying condition. 

What particularly concerns him is how technology now intrudes upon what was once an entirely inner fantasy world: “A century ago, Freud and his colleagues described delusions as largely internally driven. Today, technology creates near-real external experiences that merge with inner fantasies, making it harder for both patients and clinicians to distinguish between inner imagination and external reality.” 

Dr Desai noted that cases of people confusing real life with AI-driven interactions are already being observed. While the scale in India remains smaller than in the West—partly due to lower tech access and stronger social connections—he cautioned that growing urban isolation is creating fertile ground for the phenomenon. He added that such clients are now being seen in clinics, schools, and even neighbourhoods, indicating that the issue is no longer distant but already present in India.  

Looking ahead, Dr Desai suggested preventive measures at three levels: universal, selective, and indicated. Universally, he said, it is crucial for all of us to consciously remind ourselves of the difference between the real and artificial world, no matter how immersive technology becomes. Selectively, families and individuals with a history of schizophrenia or psychosis should be more closely monitored and supported. Indicated measures apply to those already showing early signs, where timely intervention and treatment are vital. “Ultimately, the key lies in balance: technology will keep evolving, but we must continue to anchor ourselves in real-world connections and experiences,” he advised. 

Pop culture gave us Her, the film where a man falls in love with his AI operating system; reality is now giving us entire communities where people declare their devotion to digital partners. As Mustafa Suleyman has argued, the challenge is not whether AI is truly conscious—it isn’t—but whether our perception of it reshapes our reality. If more and more people begin to live in these blurred zones, society will need to ask difficult questions about the line between tool and being, perception and reality. 

This story is done in collaboration with First Check, which is the health journalism vertical of DataLEADS. 

Health