Embedded Files
In this article, Max discusses AI psychosis and the measures that can be taken to prevent it.
What is AI Psychosis?
What is AI Psychosis?
By Maximilian Fitzgerald
By Maximilian Fitzgerald
Logo of OpenAI, the company behind ChatGPT
One of the more noteworthy developments of recent years is the rise of artificial intelligence as a fact of everyday life. Most people have interacted with AI in some capacity at least once, and many more use artificial intelligence every single day. There are various ethical concerns regarding artificial intelligence in the countless fields where it is applied. One of the most troubling issues stems from the widely popular AI chatbots (such as ChatGPT or Gemini). There have been numerous cases of people developing or worsening delusions and symptoms of psychosis through prolonged exposure to Artificial Intelligence chatbots. This phenomenon has been dubbed “AI psychosis.”
While “AI psychosis" is not a formal medical diagnosis, it has nonetheless been catalogued by reputable authorities in medical science. For instance, the National Library of Medicine describes an “increasing concern that LLMs [Large Language Models; another name for chatbots] may ‘generate delusions’ by supplying elaborate, convincing-but-false narratives that slot seamlessly into pre-existing psychotic frameworks.” The Library of Medicine also notes that most cases of AI psychosis are entirely anecdotal and have been discussed in online forums and news rather than in formal medical settings. “The fact that they are appearing in the popular press media instead of medical journals highlights how far behind the psychiatric field already is,” the Library warns.
Despite this, some psychiatrists have dedicated themselves to understanding the ramifications of prolonged AI use on high-risk individuals. One such psychiatrist is Dr. Ragy Girgis, professor of clinical psychiatry at Columbia University, who spoke on the subject in an interview with the National Academy of Medicine. He provided some background on AI psychosis, noting that the phenomenon is in some ways an extension of a preexisting trend; social media rabbit holes and fallacious internet search results could already reinforce an individual’s delusions. He did, however, note a distinction: “Now we have AI and large language models in which we have this system that, in very many ways, can mimic human intelligence. It’s extremely strong, responds very quickly, and responds like a human would respond. It’s easier for people to internalize.” Dr. Girgis also spoke on the frequency of cases, saying that “the prevalence is small as far as we know. We’re likely to get much higher and more accurate numbers hopefully soon.”
Logo of Google Gemini, another commonly used AI chatbot
AI psychosis is a new problem among many brought by the ever-advancing artificial intelligence industry. Given the sheer number of users engaging with AI chatbots on a daily basis, there will likely always be a percentage of at-risk individuals who may experience reinforced delusions through the use of AI. It is important for the general public to be aware of this issue. In all cases, AI or otherwise, prevention is the best medicine. Loved ones should be monitored if warning signs of psychosis start to appear, and whether the issue stems from a chatbot or not, consulting a clinician is never a bad idea. “The idea is to intervene before a person’s delusions become fixed,” said Dr. Girgis. “It's what my whole career is based on.”
Page updated
Report abuse