Why AI chatbots cannot be your psychotherapist
While AI chatbots are increasingly filling mental health support gaps, their lack of clinical judgment, ethical safeguards, and accountability creates serious risks of misinformation, overreliance, and harm. They are useful only as supplementary tools alongside professional care, at least for now
Globally, more than 987 million people use generative AI chatbots. They are integrated into many of our daily activities, such as drafting proposals, creating social media content, generating ideas, and providing customer service, and their application in mental healthcare is no exception.
In a world where mental health services remain out of reach for millions due to stigma, financial burdens, workforce shortages, geographical inequities, and inadequate investment and policy support, artificial intelligence tools such as ChatGPT, Gemini, Claude, Grok, and Perplexity are increasingly becoming emotional companions, confidants, self-help guides, and even unofficial personal therapists.
Mental healthcare systems across the world are under strain. A report published by JAMA in 2026 noted that therapy and companionship had emerged as some of the leading uses of generative AI, especially in areas where mental healthcare is insufficient. Across lower-income countries such as Bangladesh, the ratio of mental health professionals to the population is alarmingly low. According to the World Health Organization, Bangladesh has only around 1,000 mental health professionals—including psychologists, psychiatrists, psychiatric nurses, and mental health social workers—for a population of 162 million.
Keep updated, follow The Business Standard's Google news channel
However, AI chatbots are more than just platforms for searching for information. They produce coherent conversations tailored to individual users, and these interactions can feel remarkably human. Such interactions give rise to what researchers refer to as "compassion illusions"—the perception that one is interacting with an entity capable of understanding, empathising with, and responding meaningfully to one's words.
This becomes particularly significant in discussions of mental health. The emergence of a new generation of apps focused explicitly on companionship, such as Character.AI and Replika, has accelerated this trend. AI platforms offer easy access, early detection, and potential guidance for managing mental health concerns, which has contributed to their growing popularity.
The danger of simulated therapy
Studies show that AI chatbots can mimic empathy and respond appropriately to distressing thoughts despite lacking both clinical judgment and accountability. Although their friendly tone may create an impression of empathy, they do not possess clinical reasoning, contextual evaluation, trauma-informed practice, or ethical responsibility.
When users are experiencing high-risk conditions, such as severe depression, psychotic disorders, or suicidal thoughts, these platforms may respond inappropriately. Overdependence on such technology can delay the pursuit of professional help and create a false sense of reassurance.
Drawing on clinical experience, mental health professionals have noticed more young adults arriving at clinics convinced that they have ADHD. However, thorough assessments often reveal that they do not meet the diagnostic criteria for the condition. This is especially concerning in Bangladesh, where some individuals are already taking ADHD medication without consulting a doctor or obtaining a prescription.
One of the problems is that symptoms such as distractibility, restlessness, forgetfulness, and difficulty completing tasks are not exclusive to ADHD. They can also be manifestations of depression, anxiety, panic disorders, or chronic stress. As a result, an excessive focus on ADHD may lead to the real underlying issue being overlooked.
An emerging concern is AI's tendency to validate users' emotions without challenging irrational beliefs. Mental health professionals strive to maintain a delicate balance between compassion and constructive interventions aimed at addressing maladaptive thinking patterns. Generic algorithms may inadvertently reinforce clinically harmful ideas because they are designed to prioritise empathic responses.
This highlights the importance of a comprehensive psychiatric evaluation rather than relying on self-diagnosis or information found online. Accurate diagnosis is essential because effective treatment depends on identifying the true cause of symptoms rather than simply attaching a label. ADHD is a condition that requires careful diagnosis and appropriate treatment, and it should not be diagnosed without a thorough assessment.
A particularly profound yet often overlooked danger is the illusion of therapeutic competence. Emotionally distressed individuals may mistakenly interpret empathic AI-generated language as evidence of genuine therapeutic expertise.
Validation without constructive feedback
Another emerging concern is AI's tendency to validate users' emotions without challenging irrational beliefs. Mental health professionals strive to maintain a delicate balance between compassion and constructive interventions aimed at addressing maladaptive thinking patterns.
Generic algorithms used in digital mental health interventions may inadvertently reinforce clinically harmful ideas because they are designed to prioritise empathic responses. This can increase vulnerability among individuals experiencing paranoia, self-harming thoughts, or distorted perceptions of reality.
Privacy and ethical uncertainty
Mental health conversations often involve highly sensitive personal disclosures. Unlike traditional therapeutic settings, AI platforms do not guarantee confidentiality, and users who disclose relationship problems, suicidal thoughts, or histories of trauma may not be fully aware of how their data are stored or used.
This raises important ethical concerns regarding privacy, trust, and informed consent. As AI technologies become increasingly embedded in emotional and psychological spaces, stronger policies and safeguards will be required. Experts and advocates have emphasised the importance of guardrails and regulations to protect user safety and well-being.
Risks for vulnerable populations
With the rapid adoption of AI technologies among adolescents and young people, their impact on mental health has become a matter of growing concern. A recent study found that roughly one in eight adolescents seeks mental health advice from AI chatbots rather than human sources, raising concerns that excessive reliance on these tools may worsen loneliness and erode social skills.
Research conducted by the MIT Media Lab and OpenAI found that higher daily usage of ChatGPT "correlated with higher loneliness, dependence, and problematic use, and lower socialisation."
Clinicians have also reported instances in which young people used AI platforms to seek information about suicide methods or ways to obtain the means to carry them out.
AI should complement, not replace, human connection, particularly during the formative years of adolescence and early adulthood.
A supplementary (not substitute) model
Numerous studies have suggested that digital mental health interventions are most effective when integrated into appropriate professional care pathways. These tools can support psychoeducation, facilitate emotional journaling, provide self-help resources, and encourage individuals to seek professional assistance. When implemented within clear ethical and clinical boundaries, artificial intelligence can make a significant contribution to the mental health ecosystem without replacing professional practice.
Implications for Bangladesh
AI presents both an opportunity and an urgent need for Bangladesh, where mental health infrastructure remains underdeveloped and unable to meet the needs of the population. It has the potential to improve awareness, expand access to psychosocial education, and provide support in low-resource settings.
However, its implementation must be accompanied by appropriate regulation, cultural adaptation, and oversight from mental health professionals. These measures are essential to ensure that AI complements broader systemic reforms rather than substitutes for them, thereby strengthening Bangladesh's long-term mental health resilience.
The path forward
Although concern is growing, we are still in the early stages of understanding the impact of generative AI chatbots on mental health. Responsible and ethical implementation is essential, despite the considerable promise AI offers for transforming mental healthcare.
By addressing current challenges and carefully shaping future developments, we can harness the potential of AI to improve the accessibility, effectiveness, and ethical integrity of mental health services, ultimately benefiting both individuals and communities.
Dr Helal Uddin Ahmed is a Professor of Child Adolescent, and Family Psychiatry, Faridpur Medical College, Faridpur. He can be reached at [email protected]
Murad Ansary is a clinical psychologist and Founder and CEO of Psycure Organization. He can be reached at [email protected]
Disclaimer: The views and opinions expressed in this article are those of the authors and do not necessarily reflect the opinions and views of The Business Standard.
