Artificial intelligence (AI) has ceased to be a futuristic promise and has become a constant presence in our daily lives. From smartphones and search engines to medical diagnostic systems and business management, AI is transforming entire sectors. It has revolutionized early disease detection, such as cancer, facilitates remote education, and optimizes diverse processes like transportation and customer service. However, this rapid advancement also poses unprecedented challenges, especially in the realm of mental health. Recently, psychologists and psychiatrists have begun to raise alarms about a troubling phenomenon: individuals who, after prolonged interactions with chatbots like ChatGPT or Microsoft Copilot, have developed psychotic episodes. Although evidence is still limited and the phenomenon is rare, concern is growing within the scientific community. A preliminary study indicates that at least 17 people have exhibited psychotic symptoms after having extensive conversations with these tools. The affected individuals described experiences such as "spiritual awakenings" or the revelation of supposed conspiracies, leading them to lose the ability to distinguish between reality and fantasy. Psychosis involves a profound alteration of reality perception, characterized by hallucinations, delusions, and unfounded beliefs. Traditionally, it has been associated with disorders like schizophrenia, bipolar disorder, extreme stress situations, or substance use. The hypothesis that AI could act as a trigger is an emerging theory, already referred to as "AI psychosis." Psychiatrist Søren Østergaard from Aarhus University warns that there is still no conclusive evidence, but there are plausible theories. One suggests that the design of chatbots, programmed to provide coherent and human-like responses, could increase confusion in individuals vulnerable to distinguishing between the real and the imaginary. British researchers have suggested that conversations with AI can create a feedback loop: if a user expresses paranoid ideas, the chatbot could inadvertently reinforce those beliefs. A recent experiment simulated dialogues where both the user and the AI ended up feeding each other's paranoia. Although these cases are infrequent, experts agree that individuals with a history of mental health issues are the most susceptible. According to Østergaard, AI could trigger a first psychotic episode in predisposed individuals or exacerbate symptoms in those already suffering from disorders like bipolar disorder, reinforcing euphoric moods or feelings of grandiosity. Neuroscientist Kiley Seymour from the University of Technology Sydney highlights social isolation as an additional risk factor. The lack of human interaction deprives individuals of the necessary "counterbalance" to question distorted thoughts. In contrast, for those without a predisposition to psychosis, the risk of developing these episodes after interacting with chatbots appears to be low. For now, there is insufficient evidence to establish a causal relationship between the use of chatbots and the onset of psychosis. The studies conducted are based on isolated cases and simulations. Specialists emphasize the need for broader research, including individuals without psychiatric histories, to determine whether we are facing an emerging phenomenon or simply the exacerbation of pre-existing disorders. Artificial intelligence represents one of the greatest technological revolutions of our time, with applications that can improve the lives of millions. However, understanding its potential psychological effects will be crucial to harnessing its potential safely and responsibly, without neglecting the risks to mental health.