Kolkata : A young writer who once brimmed with enthusiasm for exploring food, culture and stories recently admitted that he could no longer write without artificial intelligence (AI) support. 

For someone who had once found joy in discovering ideas and shaping them into engaging narratives, the admission was startling. It also raised an uncomfortable question: when does a tool that makes work easier begin to undermine the very abilities it is meant to support?

Across classrooms, workplaces and personal lives, AI chatbots are becoming increasingly accessible companions for brainstorming, problem-solving, emotional reassurance and everyday decisions. For young people who have grown up communicating digitally, turning to a chatbot can feel as natural as opening a search engine. 

But the ability to generate an answer instantly can also make it tempting to outsource the difficult parts of thinking, creating and communicating.

This does not mean that using AI frequently amounts to addiction. Nor does it mean that technology inevitably diminishes creativity. The concern is what happens when people begin to feel incapable of working, making decisions or managing emotions without it.

On World Mental Health Day, experts explain why AI dependence deserves attention, how it may affect young people's psychological well-being, and why human relationships and independent thinking must remain central in an increasingly AI-driven world.

Dr Samir Parikh, Chairperson, Fortis National Mental Health Program, FMRI, Gurugram, says young people's familiarity with digital communication helps explain their early adoption of AI. "Younger people, teens, there is a lot of text conversation that they do. They are also more attuned right from their early years to go online, to search, to read online," he says.

AI chatbots build on this familiarity by offering personalised, immediate responses. Young users may also perceive these conversations as confidential and assume the answers are reliable, particularly if they have previously received correct information for academic or practical questions.

Dr Parikh explains that this trust can extend from straightforward queries to sensitive subjects, including relationships and personal struggles. "When you go to AI for a math problem," and receive a correct answer, he says, you may also begin to trust its responses when seeking relationship advice or discussing something more personal.

That shift matters because advice about emotional difficulties requires context, judgement and an understanding of an individual's circumstances. A fluent, reassuring response is not necessarily a correct or appropriate one.

When helpful AI use becomes emotional dependence
Dr Jothi Neeraja, Founder, Chairwoman and Managing Director, Maarga Mind Care, distinguishes between using AI and leaning on it emotionally.

"The concern begins when a person needs a chatbot for reassurance, validation or companionship repeatedly, and starts finding it easier than speaking to people," she explains.

Dependence becomes more concerning when it begins affecting relationships, sleep, work or independent decision-making. Someone may repeatedly ask a chatbot what to think, feel or do, gradually becoming less confident about making choices without its input.

AI's constant availability can reinforce this pattern. Unlike human relationships, chatbot interactions do not ordinarily involve waiting for a response, navigating disagreement or worrying about tiring someone out. For people experiencing loneliness or social anxiety, this can offer immediate relief.

However, Dr Neeraja warns that comfort can gradually turn into dependence. Human relationships involve vulnerability, differences of opinion and the possibility of rejection. Learning to navigate these experiences is also part of developing emotional resilience and the ability to connect with others.

Emotional dependence is only one part of the concern. The young writer's experience raises another question: what happens when people repeatedly outsource the process of generating ideas, structuring arguments or expressing themselves?

AI can be a useful thinking partner. It can help writers overcome a blank page, organise research, identify gaps in an argument and explore alternative perspectives. But if users routinely accept generated responses without questioning, revising or developing them independently, they may lose opportunities to practise those skills.

The distinction is between using AI to extend one's abilities and using it to avoid the effort through which those abilities develop.

For students and early-career professionals, this is particularly important. Writing is not simply the production of polished sentences; it also involves discovering what one thinks, weighing evidence, making connections and developing an individual voice. When AI does most of that work, the final output may improve while the user's own learning remains limited.

This is a concern about habitual reliance, not proof that AI use inevitably damages the brain or permanently reduces creativity. The long-term effects of different patterns of AI use still require careful study.

Warning signs families and young people should watch for 
Dr Neeraja says the most useful indicators are changes in everyday functioning rather than the number of chatbot conversations alone.

Finding it increasingly difficult to tolerate uncertainty or manage emotions without chatbot reassurance
Another concern is the reinforcement of unhealthy beliefs. If a vulnerable person repeatedly receives validation for paranoid thoughts, self-harm ideas or harmful relationship patterns, it may become harder to question those beliefs or seek appropriate help.

Dr Neeraja cautions that mental health concerns require assessment and human judgement, neither of which should be replaced by a chatbot.

India's large, young digital population and uneven access to mental health professionals make this an important conversation. AI can offer a convenient starting point for people who feel embarrassed about seeking help, cannot readily access a specialist or do not know where to begin.

But accessibility should not be confused with clinical reliability. A chatbot cannot be assumed to provide a diagnosis, assess risk accurately or offer appropriate treatment simply because its response sounds empathetic.

Dr Parikh says the response must involve education across society, including the media, role models, mental health professionals and educational institutions. Schools and colleges need to help young people understand both the benefits and limitations of AI, while parents should encourage open conversations rather than shame or punish children for using it.

He also highlights the responsibility of platforms to make professional help prominent when users discuss distress or other sensitive issues.

Dr Neeraja suggests asking what an emotionally dependent user is seeking from AI: companionship, reassurance, a listener or freedom from judgement. Understanding that unmet need is more useful than simply telling someone to stop using a chatbot.

AI can support learning, creativity and access to information when used deliberately. The safeguard is to keep practising independent thinking, maintain relationships outside the screen and seek qualified help when emotional difficulties begin affecting daily life. As Dr Neeraja puts it, "The aim should be to use AI to connect people with support, not to make people feel that they no longer need people."

The challenge for India is not to reject AI, but to ensure that a generation growing up with it does not mistake instant answers for wisdom, simulated companionship for human connection, or assistance for the ability to think independently.