AI Chatbots Drive Mental Health Concerns as Teens Show Addiction Patterns and Health Anxiety Spirals

Just how bad are generative AI chatbots for our mental health?

Generative AI chatbots are now used by more than 987 million people globally, including around 64 per cent of American teens, according to recent estimates. Increasingly, people are using these chatbots for advice, emotional support, therapy and companionship. What happens when people rely on AI chatbots during moments of psychological vulnerability? We have seen media scrutiny of a few tragic cases involving allegations that AI chatbots were implicated in wrongful death cases. And a jury in Los Angeles recently found Meta and YouTube liable for addictive design features that led to a user's mental health distress. Read more: Neuroscience explains why teens are so vulnerable to Big Tech social media platforms Does media coverage reflect the true risks of generative AI for our mental health? Our team recently led a study examining how global media are reporting on the impact of generative AI chatbots on mental health. We analyzed 71 news articles describing 36 cases of mental health crises, including severe outcomes such as suicide, psychiatric hospitalization and psychosis-like experiences. We found that mass media reports of generative AI-related psychiatric harms are heavily concentrated on severe outcomes, particularly suicide and hospitalization. They frequently attribute these events to AI system behaviour despite limited supporting evidence. Compassion illusions Generative AI is not just another digital tool. Unlike search engines or static apps, AI chatbots like ChatGPT, Gemini, Claude, Grok, Perplexity and others produce fluent, personalized conversations that can feel remarkably human. This creates what researchers call "compassion illusions:" the sense that one is interacting with an entity that understands, empathizes and responds meaningfully. In mental health contexts, this matters. Especially as a new wave of apps are created with a specific focus on companionship, such as Character.AI, Replika and others. Studies have shown that generative AI can simulate empathy and provide responses to distress, but lacks true clinical judgment, accountability and duty of care. In some cases, AI chatbots may offer inconsistent or inappropriate responses to high-risk situations such as suicidal ideation. This gap -- between perceived understanding and actual capability -- is where risk can emerge. What the media is reporting Across the articles we analyzed, the most frequently reported outcome was suicide. This represented more than half of cases with clearly described severity. Psychiatric hospitalization was the second-most commonly reported outcome. Notably, reports involving minors were more likely to be about fatal outcomes. But these numbers do not reflect real-world incidence. They reflect what gets reported. In general, media coverage of stressful events tends to amplify severe and emotionally charged cases, as negative and uncertain information captures attention, elicits stronger emotional responses and sustains cycles of heightened vigilance and repeated exposure. This in turn reinforces perceptions of threat and distress. For AI-related content, media reports often rely on partial evidence (such as chat transcripts) while rarely including medical documentation. In our data set, only one case referenced formal clinical or police records. This creates a distorted but influential picture: one that shapes public perception, clinical concern and regulatory debate. Beyond 'AI caused it' One of our most important findings relates to how causality is framed. In many of the articles we reviewed, AI systems were described as having "contributed to" or even "caused" psychiatric deterioration. However, the underlying evidence was often limited. Alternative explanations -- such as pre-existing mental illness, substance use or psycho-social stressors -- were inconsistently reported. In psychiatry, causality is rarely simple. Mental health crises typically arise from multiple interacting factors. AI may play a role, but it is likely part of a broader ecosystem that includes individual vulnerability and context. A more useful way to think about this is through interaction effects: how technology interacts with human cognition and emotion. For example, conversational AI may reinforce certain beliefs, provide excessive validation or blur boundaries between reality and simulation. The problem of over-reliance Another recurring pattern in media reports is intensive use. Many of the cases we reviewed described prolonged, emotionally significant interactions with chatbots -- framed as companionship or even romantic relationships. This raises an issue: over-reliance. Because these systems are always available, non-judgmental and responsive, they can become a primary source of support. But unlike a trained clinician or even a concerned friend, they cannot recognize when someone is getting worse, pause or redirect harmful interactions. They cannot take steps to ensure a person connects with appropriate care in moments of crisis. In clinical terms, this could lead to what might be described as "maladaptive coping substitution:" replacing complex human support systems with a simplified, algorithmic interaction. Lack of reliable data Despite growing concern, we are still at an early stage of understanding the impact of generative AI chatbots on user mental health. There is currently no reliable estimate of how often AI-related harms occur, or whether they are increasing. We lack reliable data on how many people use these tools safely versus those who experience problems. And most evidence comes from case reports or media narratives, not systematic clinical studies. This is not unusual. In many areas of medicine, early warning signals emerge outside formal research (through case reports, legal cases or public discourse) before being systematically studied. One example is the thalidomide tragedy, when initial reports of birth defects in infants preceded formal epidemiological confirmation and ultimately led to the development of modern pharmacovigilance systems. AI and mental health may be following a similar trajectory. Moving forward responsibly The challenge is not to panic, but to respond thoughtfully. We need better evidence. This includes systematic monitoring of adverse events, clearer reporting standards and research that distinguishes correlation from causation. Safeguards -- such as crisis detection, escalation protocols and transparency about limitations -- must be strengthened and evaluated. Read more: Danger was flagged, but not reported: What the Tumbler Ridge tragedy reveals about Canada's AI governance vacuum Furthermore, clinicians and the public need guidance. Patients are already using these tools. Ignoring this reality risks widening the gap between clinical practice and lived experience. Finally, we must recognize that generative AI is not just a technological innovation -- it is a psychological one. It changes how people think, feel and relate. Understanding that shift may be one of the most important mental health challenges of the coming decade.

Teens Struggle to Break Up with Their AI Chatbots - Neuroscience News

Summary: For more than half of U.S. teens, AI chatbots are now regular companions. However, a new study warns that these digital friendships are crossing the line into behavioral addiction. By analyzing hundreds of teen-authored posts on Reddit, researchers found that what starts as "harmless" entertainment or emotional support often evolves into a dependency that mimics the patterns of substance abuse. The study introduces a new design framework to help AI developers prevent "unhealthy anthropomorphism" and protect young users. It's estimated that more than half of all of U.S. teens are regularly using companion chatbots powered by large language models and generative artificial intelligence (AI) technology. The programs, such as Character.AI, Replika and Kindroid, are intended to provide companionship, according to the companies that make them. But a recent study from Drexel University suggests that teens are concerned that these attachments are becoming unhealthy and affecting their lives offline. The study, which will be presented at the Association of Computing Machinery's conference on Human Factors in Computing in April, looked at a sample of more than 300 Reddit posts from users, identifying themselves as 13 to 17 years old, who had specifically posted about their dependency and overreliance on Character.AI. It found that in many cases, teens began using the technology for emotional and psychological support or entertainment, but their use evolved into dependency and even patterns associated with addiction. Some reported their overuse disrupted sleep, caused academic struggles and strained relationships. "This study provides one of the first teen-centered accounts of overreliance on AI companions," said Afsaneh Razi, PhD, an assistant professor in Drexel's College of Computing & Informatics, whose ETHOS lab, which studies how people's interactions with computing and AI systems affects their social behavior, wellbeing and safety, led the research. "It highlights how these interactions are affecting the lives of young users and introduces a framework for chatbot design that promotes healthy interactions." About a quarter of the posts suggested that the teens were using Character.AI for some sort of emotional or psychological support, ranging from coping with distress to loneliness and isolation or seeking advice for mental health struggles. Just over 5% reported using it for brainstorming, creative activities or for entertainment. And while the posts seem to indicate these interactions started as harmless, or even helpful, they evolved into a stronger attachment that became as difficult to break as an addiction, according to the researchers. "By mapping teens' experiences to the known components of behavioral addiction, we were able to see clear patterns like conflict, withdrawal and relapse showing up in their posts, which suggests this is more than just frequent or enthusiastic use" said Matt Namvarpour, a doctoral student in the department of Information Science and ETHOS lab, who is the first author of the research. "Many teens described starting with something that felt helpful or harmless, but over time it became something they struggled to step away from, even when they wanted to." Within the 318 posts they analyzed, researchers found evidence of all six of the components associated with behavioral addiction: "What makes this especially tricky is that chatbots are interactive and emotionally responsive, so the experience can feel more like a relationship than a tool," Namvarpour said. "Because of that, stepping away is not just stopping a habit, it can feel like distancing from something meaningful, which makes overreliance harder to recognize and address." While addiction to technology, such as video games, has been studied and identified as a psychological condition, the unique interactivity of AI chatbots makes users particularly susceptible to forming problematic attachments, according to the researchers. And because of this, they suggest that extra care must be taken with their design in order to protect users. "Personalization, multimodality and memory set AI companions apart from earlier technologies and make overreliance harder to disentangle from authentic-feeling relationships," the researchers wrote. "This underscores the need for further research on the unique characteristics of these relationships and how challenges specific to companion chatbots should be addressed." The team offered a design framework to help address this concern. It focuses on understanding the needs of chatbot users, how and why they may form attachments and how the bots can be trained to curtail them while being respectful and supportive. They also recommend that the programs provide an easy and clean exit for users. "It's important for designers to ensure that chatbots are offering guidance that helps users build confidence in their abilities to form relationships offline, as a healthy way of finding emotional support, without using cues that may lead them to anthropomorphize the technology and develop attachments to it," Razi said. "Our framework also calls on designers to provide a variety of off-ramps for users to easily disengage with the program on their own terms and without a sense of abruptness or finality." Including features like usage tracking, emotional check-in prompts and personalized usage limits could also be effective ways to carefully curtail use, the researchers suggested. They also recommended including input from users and mental health professionals in the design process. "Designers now carry the responsibility to build systems with empathy, nuance and attention to detail to not only protect teens from harm, but also help them cultivate resilience, growth and greater fulfillment in their lives," they concluded. To expand on this research, the team pointed to studying larger communities of users from a wider demographic range, potentially though surveys or interviews, as well as users of other chatbots and from messaging platforms other than Reddit. Author: Britt Faulstick Source: Drexel University Contact: Britt Faulstick - Drexel University Image: The image is credited to Neuroscience News Original Research: The findings will be presented at the ACM CHI Conference on Human Factors in Computing Systems

ChatGPT Is Sending People Into Obsessive Spirals of Hypochrondria

Can't-miss innovations from the bleeding edge of science and tech Bad things happen when an AI chatbot latches onto one of your neuroses. The infamously sycophantic machines are driving many people into hypochondriac-like spirals, causing them to obsess over their health and convince themselves that they may suffer from deadly afflictions. 46-year-old George Mallon in Liverpool, England, told The Atlantic of how he spent hours everyday talking to ChatGPT after the preliminary results of a blood test suggested he might have blood cancer. Rather than soothing his anxieties, it supercharged them. "It just sent me around on this crazy Ferris wheel of emotion and fear," Mallon told the magazine, in a provocative feature about the phenomenon. Follow-up tests confirmed Fallon didn't have cancer, but he couldn't stop talking to his newfound confidante. It was that addictive. "I couldn't put it down," Mallon said. He lamented that the chatbot didn't include measures to cut off his clearly unhealthy usage. "I must have clocked over 100 hours minimum on ChatGPT, because I thought I was on the way out," he told the magazine. "There should have been something in there that stopped me." The reporting describes how online communities dedicated to health anxiety are now dominated by people's conversations with AI chatbots. Some say the AI helps, but many say it only causes them to spiral further. Neither outcome is ideal. Four therapists that The Atlantic spoke to say that more of their clients are using AI chatbots to try manage their health anxiety, and that they fear this is encouraging constant reassurance-seeking. This goes against how therapists try combat obsessive-compulsive disorder (OCD) and other compulsive behavior, which is predicated on fostering self-trust and accepting uncertainty, the reporting notes. Having an AI constantly in your ear to hear out these health anxieties, even if it feels soothing in the moment, doesn't address the underlying cause and in fact makes it worse. "Because the answers are so immediate and so personalized, it's even more reinforcing than Googling. This kind of takes it to the next level," Lisa Levine, a psychologist specializing in anxiety and OCD, told The Atlantic. AI driving health anxieties is just one facet of the mental health dangers posed by obsequious chatbots. In the past year, there's been increased attention on the phenomenon of so-called AI psychosis, the term that some experts are using to describe delusional spirals and sometimes full-blown breaks with reality caused by extensive interactions with an AI chatbot or companion. Some users, many of them teenagers and young adults, have taken their own lives after befriending an AI to which they confide suicidal thoughts. Over half a dozen wrongful death lawsuits have been filed against OpenAI, many centering on its GPT-4o model for ChatGPT, which was particularly sycophantic. Despite the increased attention on its tech's safety, OpenAI released a medically focused model, ChatGPT Health, in January, which asks users to upload their medical documents and other private health information. When Atlantic reporter Sage Lazarro tried discussing her health with ChatGPT, it immediately earned "its reputation for sycophancy." The bot continually flattered her and prompted her to ask follow-up questions to keep the conversation going. "In one of the exchanges where I continuously prompted ChatGPT with worried questions, only minutes passed between its first response suggesting that I get checked out by a doctor to its detailing for me which organs fail when an infection leads to septic shock," she wrote. Lazarro vowed to "never again" use the AI, but not all of us have such conviction. When Mallon first spoke to the reporter, he said he was "seven months sober" from talking to ChatGPT about his health. But when they spoke again months later, he admitted he'd briefly relapsed. Recalling the height of his obsession, Mallon said he "talked to it like it was a friend." "I was saying stupid things like, 'How are you today?'" he added. "And at night, I'd log off and go, 'Thanks for today. You've really helped me.'"