Young people ditch AI chatbots and return to human support through mental health service in India
Mental health has become a central topic in tech conversations over the past few years. With the rise of tools like ChatGPT, a lot of people started using AI as a kind of makeshift emotional support system, especially young people who couldn’t afford a private therapy session. It makes sense, right? Easy access, available around the clock and, of course, free.
But what’s happening in India tells a different story, and a pretty interesting one at that. A growing number of young people are ditching chatbots and going back to calling Tele-MANAS, the Indian government’s psychological counseling service. And the reasons behind this shift say a lot about what AI still can’t do, at least when it comes to truly caring for someone who is suffering.
The service currently handles between 120 and 130 calls per day, and about 60% to 70% of those calls are related to relationship problems, breakups, academic stress and emotional distress among students and young adults. That volume reveals not just the demand for support, but also the intensity of what this generation faces on a daily basis.
One of the accounts that caught the attention of the service’s counselors sums up the problem pretty well:
The chatbot told me I was overreacting, but I still felt like hurting myself.
Another young person shared: I think I confused the system. It didn’t understand what I was trying to say.
Those statements alone raise a huge question about the limits of artificial intelligence when it comes face to face with the complexity of human emotions. 🤔
What is Tele-MANAS and why does it matter
Tele-MANAS, which stands for Mental Health and Normalcy Augmentation System, is a national mental health program launched by the Indian government in 2022. It works as a free psychological support hotline, accessible by phone, that connects anyone with trained professionals like psychologists and clinical counselors in real time. The idea has always been simple: democratize access to emotional care in a country with over 1.4 billion people and a chronic shortage of mental health professionals available through the public system.
For a long time, the service faced the classic challenge of any public health policy: how do you reach the people who need it most? Young people, in particular, rarely seek formal help. The combination of social stigma, lack of information and, most importantly, the financial barrier to private appointments meant that many simply never asked for help. It was precisely in that gap that AI chatbots found room to grow as a makeshift emotional alternative. And it worked for a while, at least on the surface.
Receive the best innovation content in your email.
All the news, tips, trends, and resources you're looking for, delivered to your inbox.
By subscribing to the newsletter, you agree to receive communications from Método Viral. We are committed to always protecting and respecting your privacy.
What the latest data from Tele-MANAS shows, however, is a reversal nobody expected so soon. The volume of calls from young people has increased steadily, and a significant portion of those users reported having tried an AI chatbot before calling. Many described their experience with AI as frustrating, cold or even harmful, and returned to seeking human interaction as a way to feel truly heard.
The cost of therapy pushes young people toward AI
One factor you really can’t ignore in this story is money. A student interviewed by the Times of India explained the situation pretty directly: he tried to seek private professional counseling, but a single session cost 1,500 rupees or more, roughly equivalent to about $18. For many Indian college students, that amount is simply unaffordable, especially when treatment requires ongoing follow-up with multiple sessions.
Unable to cover the cost, this student turned to AI as a therapeutic alternative. And he’s far from alone. Thousands of young people around the world do the same thing every day, opening ChatGPT or similar tools to vent, ask for advice about anxiety, deal with panic attacks or simply find someone, even an artificial one, who seems to be listening.
The problem is that this apparent accessibility hides some serious pitfalls. When the cost of professional therapy is prohibitive and the most accessible free alternative is a chatbot with no real capacity for emotional understanding, it creates a dangerous gap in mental health care. And that’s exactly the gap that programs like Tele-MANAS are trying to fill by offering free care with real professionals.
Why AI fails when it comes to real emotion
There is a massive difference between processing language and understanding suffering. Large language models, the well-known LLMs like GPT-4 and similar tools, are incredible at organizing information, summarizing text, answering questions and even simulating conversations that feel natural. But when someone shows up carrying a real emotional burden, the response generated by an algorithm, no matter how sophisticated, still carries a structural limitation: it’s based on statistical text patterns, not genuine empathy, not active listening and definitely not emotional intelligence in the human sense of the word.
Dr. Jawaharlal Nehru P, a senior psychologist at Tele-MANAS, explained this issue clearly. According to him, AI has a very limited understanding of emotions. It mirrors what you say, but it can’t truly grasp the feeling behind the words. It’s not capable of guiding behavior in a meaningful or personalized way.
He also added that mental health issues require a deeper assessment. Everything that affects behavior needs proper analysis, and AI can’t provide a balanced psychological evaluation or understand context beyond what is typed.
Dr. Vivaswan Boorla, a senior psychologist at the Institute of Mental Health in Hyderabad, reinforced this view. He pointed out that while AI can offer structured responses, it lacks the ability to interpret human emotions and pick up on subtle cues that make all the difference in a counseling session. In his words, the attunement of human interaction is completely different. What you receive from a trained professional, like empathy, warmth and understanding, cannot be matched by a machine.
The case reported by Tele-MANAS counselors illustrates this point well. When a young person in distress shares thoughts of self-harm and gets back a response that minimizes what they’re feeling, telling them they’re overreacting, the effect can be devastating. Not because the AI intended to do that, it obviously has no intention, but because the model simply doesn’t have the ability to calibrate the real weight of those words within the context of an emotional crisis. It lacks what psychologists call therapeutic presence: the perception that there is someone on the other side who genuinely cares about what happens to you.
This doesn’t mean artificial intelligence is useless in the mental health space. There are very valid applications, like initial symptom screening, providing information about disorders, medication reminders or complementary support between appointments. But placing a chatbot in the role of a mental health professional’s substitute, especially during moments of crisis, is a real risk that is now being discussed with more seriousness precisely because of stories like those reaching Tele-MANAS every day. 😔
Active listening and the power of your native language
One detail that often goes unnoticed in this discussion is the role of language. The specialists at Tele-MANAS highlight that one of the service’s greatest advantages is allowing people to express themselves freely in their native language. In India, where there are dozens of languages and regional dialects, this makes a huge difference. When someone can describe what they feel in the language they think and dream in, emotional communication flows in a much more authentic and precise way.
AI chatbots, while increasingly multilingual, still show uneven performance across languages. Most models were predominantly trained in English, which means interactions in other languages can lose nuance, cultural context and even generate responses that simply don’t make sense within a specific local reality.
Tele-MANAS counselors also stress that their work goes far beyond giving advice. They practice active listening, build trust throughout the conversation and offer a safe space where people feel comfortable opening up. This process of building connection, even if brief, is something that current technology cannot replicate.
As Dr. Nehru himself pointed out, sometimes people don’t want immediate treatment. They just want someone who will listen. That sentence probably sums up what’s at stake in this discussion better than any technical analysis ever could.
Relationship conflicts dominate the calls
The mental health professionals at the service have observed that a significant share of calls involves emotional conflicts tied to relationships. These are situations marked by unmet expectations, communication problems between couples, painful breakups and the difficulty of dealing with rejection.
This type of issue is especially complex because it involves multiple layers of feeling, personal history and social context that would be nearly impossible for an automated system to capture. A chatbot can suggest generic coping techniques, but it can’t tell when a person is minimizing their own pain, when there’s a concerning behavioral pattern behind the words or when the situation requires a referral to a psychiatrist.
The specialists emphasize that psychiatric care is not just about offering solutions. It’s about understanding the depth of personal suffering, providing non-judgmental support and, often, simply being present while the person finds their own answers.
What this shift tells us about human interaction and technology
Young people returning to Tele-MANAS is not a rejection of technology. It’s actually a much more nuanced signal than it appears at first glance. These same young people grew up with a smartphone in hand, use social media fluently, try out new apps without hesitation and yet, when it comes to mental health, they chose to call a human being. That says a lot about what people truly need when they’re vulnerable: real connection, not simulated connection.
Human interaction has qualities that no language model can fully replicate, at least not with current technology. A trained counselor notices the silence between words, adjusts their tone of voice to match the emotion of the moment, pauses when necessary and conveys, even over the phone, the feeling that there is someone present and committed to the well-being of the person on the other end of the line. This quality of listening is not just a technical advantage. It’s the core of what makes psychological support effective, especially in crisis situations.
Technology as a complement, not a replacement
The experts interviewed for this story are not against artificial intelligence. They acknowledge that AI can play a supporting role as a first step for basic guidance. The problem shows up when the tool is used as a complete substitute for professional care, something it was simply never designed for.
As Dr. Boorla put it, the ideal approach is to use technology as a supplement, not a replacement. This perspective is increasingly shared by mental health professionals around the world. AI can help identify early symptoms, suggest breathing exercises, provide psychoeducational information and even facilitate someone’s first contact with the healthcare system. But clinical decision-making, emotional support and building a therapeutic plan remain tasks that require the presence and judgment of a trained human being.
AI tools operate within programmed boundaries and do not possess independent thought or emotional awareness. Recognizing that boundary is not diminishing technology. It’s using it in a smarter and safer way.
The movement happening in India could be an important indicator for the rest of the world. As more people experience the limits of chatbots in emotional contexts, the conversation about AI’s role in mental health is likely to gain more nuance and responsibility. It’s not about choosing between technology and humanity, but about understanding where each one truly serves a purpose. And for now, when the suffering is real, the human voice on the other end of the line still makes a difference that no algorithm has managed to replicate. 💙
What the Tele-MANAS case makes clear is that emotional intelligence is not a feature that can be programmed, at least not in the way people need it when they’re at their breaking point. And recognizing that limit is perhaps the first step toward using technology in a more responsible and humane way.
