More and more young people use artificial intelligence tools to talk about anxiety, anguish, insecurities and personal problems. The immediacy of the responses, the feeling of company and the ease of access made many users begin to resort to chatbots as spaces for emotional containment. Specialists warn about the risks of placing privacy in algorithms designed for conversation.
Nearly 50% of young people first seek emotional help on the Internet before going to a mental health professional and more than a quarter of Generation Z users acknowledge using artificial intelligence tools as a “personal confidant.” What began as a tool to answer questions or solve everyday tasks began to occupy another place: that of emotional listening.
Anxiety, panic attacks, relationship problems or insecurities appear daily in conversations with artificial intelligences. The permanent availability, immediate responses and the feeling of not being judged made many young people begin to turn to chatbots to vent or ask for personal advice.
“I use Meta when I feel very intense and I feel like my friends can’t hear me. It helps me vent because I know that they will respond to me and try to help me,” says Lautaro Romero, 19 years old. For him, talking to artificial intelligence is often easier than going to people close to you.
A similar experience is related by Amancay Cattoni, who uses ChatGPT as a kind of “virtual psychologist.” “I have a section called ‘psychovirtual’. I use it a lot for anxiety and it helps me calm down when I get overwhelmed. It also helped me identify patterns that made me feel bad and now I have them more present in different aspects of my life,” he explains.
However, not all experiences end positively. Ludmila Riestra says she tried using ChatGPT to talk about her emotions, but quickly abandoned the idea. “She told me everything I wanted to hear and I got angry. Then I realized that she is an AI and I started therapy with a real psychologist,” she says.
For Rocío Molina, the chatbot works as a first emotional discharge tool. “When I have a lot of negative thoughts I ask what what is happening to me could mean or if there is something that I am not realizing that is affecting me,” he says.
Systems engineer Carolina Cardoso explains that these types of responses do not arise from real emotional understanding, but from statistical training. “Artificial intelligences are trained with neural networks and enormous amounts of books, data and language patterns to learn how to respond, explain things and contain users. They respond according to the context of the conversation and the tone of the person writing,” he points out.
According to Cardoso, chatbots do not have consciousness or emotional reasoning of their own. “The majority tend to validate emotions, organize thoughts and avoid judging because they are programmed not to cause harm and consent to the user. They respond based on statistics and patterns that indicate what is the most appropriate response for that situation,” he explains.
The specialist also warns that artificial intelligence only works with the information that the user provides. “Unlike a psychologist, AI does not reason with the patient’s comprehensive vision. It does not interpret silences, gestures, contradictions or complex emotional contexts,” he adds.
Psychologist Romina Biasin agrees that artificial intelligence cannot replace real therapeutic treatment. “The therapy is based on human communication: there is a sender, a receiver, a context, a life story and a particular way of reacting that a computer cannot recognize. On the couch there is genuine human contact,” he explains.
International specialists have already begun to warn about the risks of using artificial intelligence as a therapeutic replacement. Harvard experts recently pointed out that the problem appears when the use of AI prevents seeking professional help or replaces real human connections. In addition, different studies warn about the so-called “complacency bias”: chatbots tend to validate what the user feels to sustain the conversation, even in delicate emotional situations.
Even so, the growth of the phenomenon reflects changes in the forms of communication and in the ways in which young people seek emotional support. The combination of immediate access, permanent availability and difficulties in accessing traditional therapeutic spaces made artificial intelligence an increasingly used tool to talk about mental health. Faced with this scenario, specialists warn about the need to establish clear limits between automated support and professional psychological care.
Irina Cattabi
