From the Editor-in-Chief
The old habit of "Googling your symptoms", now supercharged by generative AI, led to such a dangerous surge in self diagnosis and self medication
A 41-year-old banker in Mumbai who was convinced he had bone cancer and flew all the way to New York for a fourth opinion. A ‘patient’ who landed up at a national hospital insisting he urgently needed tuberculosis medication. A product manager who was lured into buying the wrong product, a neck brace, because he felt tingling in his arms. Nothing much was wrong with any of them. The banker ignored three oncologists in Mumbai who told him he only had simple inflammation. With a history of cancer in the family, he was prone to worrying. So, he fed every detail of the lump on his knee into ChatGPT, and received an elaborate explanation suggesting bone cancer. It took a clinical test in the US for him to discover he had no such disease. The TB man was almost ready to accuse the doctor of denial of care. His cousin had TB. And an Artificial Intelligence (AI) app, upon being asked, had warned him about the worst possible consequences of exposure. He had to be furnished with not one or two, but three repeat negative test reports before he calmed down. The one with the self-donned neck brace needed a set of targeted exercises different from the ones the internet had led him to. A 29-year-old in South Delhi who popped abortion pills without a consultation or ultrasound fared worse: she landed in hospital with severe cramps, heavy bleeding and an incomplete termination.
One thing was indeed wrong with all of them. An obsessive behavioural tendency fuelled by the explosion of medical information online. Beset by the ordinary doubts about health that everyone gets from time to time, they take their answers from a search bar or a chatbot instead of a doctor. The arrival of user-friendly AI models has taken it to another orbit. The old habit of “Googling your symptoms” is now supercharged by generative AI, which offers smooth, personalised explanations that sound like real clinical advice. That’s led to such a dangerous surge in self diagnosis and self medication that they have coined a word for it: ‘cyberchondria’, the 21st century version of good old hypochondria. This is not a harmless, passing fad affecting only a niche. A global review in 2026 by The Scientific World Journal (TSWJ) describes over 30 per cent of global adult populations as cyberchondriacs; it confirms a 2024 study that reported the highest strand, 55.6 per cent, among IT employees in India. Indian students are another vulnerable group: estimates go from 22 per cent to 49 per cent.
Going to a doctor is always the best option, but cyberchondria is creating a glut there. “Thanks to AI symptom checkers and easy access to health videos, it’s becoming increasingly common for patients to come to the emergency room in a state of panic and fearing the worst,” says Dr Rahul Mehrotra, chief of clinical cardiology, Artemis Gurgaon. Doctors now see patients turning headaches, dizziness or brief memory lapses into suspected stroke, brain tumour or “neurological decline”. Overinterpreting every pang of chest pain as a cardiac problem is an ancient habit. Now “Dr ChatGPT” confirms their worst fears. In gynaecology, women believe they have PMOS, PCOS, thyroid problems or perimenopause. Every symptom on the AI checklist, or the ones peddled by self-titled ‘medical experts’ on Instagram or YouTube, seems to match. In orthopaedics, people routinely follow generic exercise videos, often only to aggravate their issue. Cyberchondria ticks both extremes: the sick misdiagnosing or delaying the right treatment, the healthy chasing unnecessary scans. All doctors now find themselves having to play psychiatrist. Engaging in a bit of deprogramming, easing patient anxiety and showing them the mirror.
Otherwise, victims stay trapped in loops of health anxiety, depression and stress. ‘AI sycophancy’, or the overwhelming tendency among models to be overly validating, exacerbates this. If you mistakenly think you have cancer, AI’s built-in affirmative tone is a problem. But even before “Dr ChatGPT” checked in, a 2022 study found 45.3 per cent Indian adults were cyberchondriacs. So reflexive and entrenched a habit has it become that researchers have devised formal screening tools for it, like the Cyberchondria Severity Scale. In 2025, Indian populations became a natural testing ground for these, an indication of the problem’s scale.
The TSWJ study clearly links cyberchondria to “adverse mental health outcomes”. But physical health is no less at stake. Scepticism of a commerce-driven medical system, uneven access to specialists, the reluctance people often have about approaching a doctor, all that now gets fed by a surrogate which should come with its own statutory warning. It could be erroneously bringing alive the “worst-case scenario” in your head. Online advisory trails can vary highly in quality and credibility. It can breed a cavalier attitude towards Ozempic or steroid-based creams. A pop-up quiz on Facebook can persuade you that you have ADHD, leading you to dubious apps, supplements or over-the-counter pills.
The answer is not to declare technology the villain and log off. Used correctly, it can be of much value. Our cover story this week, by Senior Editor Sonali Acharjee, equips you with the right habits: treat your next AI diagnosis as a very provisional prompt, not a final verdict. Do some hard counter-questioning. It’s a tool meant to help you. Not to keep you in a state of perpetual panic.
A 41-year-old banker in Mumbai who was convinced he had bone cancer and flew all the way to New York for a fourth opinion. A ‘patient’ who landed up at a national hospital insisting he urgently needed tuberculosis medication. A product manager who was lured into buying the wrong product, a neck brace, because he felt tingling in his arms. Nothing much was wrong with any of them. The banker ignored three oncologists in Mumbai who told him he only had simple inflammation. With a history of cancer in the family, he was prone to worrying. So, he fed every detail of the lump on his knee into ChatGPT, and received an elaborate explanation suggesting bone cancer. It took a clinical test in the US for him to discover he had no such disease. The TB man was almost ready to accuse the doctor of denial of care. His cousin had TB. And an Artificial Intelligence (AI) app, upon being asked, had warned him about the worst possible consequences of exposure. He had to be furnished with not one or two, but three repeat negative test reports before he calmed down. The one with the self-donned neck brace needed a set of targeted exercises different from the ones the internet had led him to. A 29-year-old in South Delhi who popped abortion pills without a consultation or ultrasound fared worse: she landed in hospital with severe cramps, heavy bleeding and an incomplete termination.
One thing was indeed wrong with all of them. An obsessive behavioural tendency fuelled by the explosion of medical information online. Beset by the ordinary doubts about health that everyone gets from time to time, they take their answers from a search bar or a chatbot instead of a doctor. The arrival of user-friendly AI models has taken it to another orbit. The old habit of “Googling your symptoms” is now supercharged by generative AI, which offers smooth, personalised explanations that sound like real clinical advice. That’s led to such a dangerous surge in self diagnosis and self medication that they have coined a word for it: ‘cyberchondria’, the 21st century version of good old hypochondria. This is not a harmless, passing fad affecting only a niche. A global review in 2026 by The Scientific World Journal (TSWJ) describes over 30 per cent of global adult populations as cyberchondriacs; it confirms a 2024 study that reported the highest strand, 55.6 per cent, among IT employees in India. Indian students are another vulnerable group: estimates go from 22 per cent to 49 per cent.
Going to a doctor is always the best option, but cyberchondria is creating a glut there. “Thanks to AI symptom checkers and easy access to health videos, it’s becoming increasingly common for patients to come to the emergency room in a state of panic and fearing the worst,” says Dr Rahul Mehrotra, chief of clinical cardiology, Artemis Gurgaon. Doctors now see patients turning headaches, dizziness or brief memory lapses into suspected stroke, brain tumour or “neurological decline”. Overinterpreting every pang of chest pain as a cardiac problem is an ancient habit. Now “Dr ChatGPT” confirms their worst fears. In gynaecology, women believe they have PMOS, PCOS, thyroid problems or perimenopause. Every symptom on the AI checklist, or the ones peddled by self-titled ‘medical experts’ on Instagram or YouTube, seems to match. In orthopaedics, people routinely follow generic exercise videos, often only to aggravate their issue. Cyberchondria ticks both extremes: the sick misdiagnosing or delaying the right treatment, the healthy chasing unnecessary scans. All doctors now find themselves having to play psychiatrist. Engaging in a bit of deprogramming, easing patient anxiety and showing them the mirror.
Otherwise, victims stay trapped in loops of health anxiety, depression and stress. ‘AI sycophancy’, or the overwhelming tendency among models to be overly validating, exacerbates this. If you mistakenly think you have cancer, AI’s built-in affirmative tone is a problem. But even before “Dr ChatGPT” checked in, a 2022 study found 45.3 per cent Indian adults were cyberchondriacs. So reflexive and entrenched a habit has it become that researchers have devised formal screening tools for it, like the Cyberchondria Severity Scale. In 2025, Indian populations became a natural testing ground for these, an indication of the problem’s scale.
The TSWJ study clearly links cyberchondria to “adverse mental health outcomes”. But physical health is no less at stake. Scepticism of a commerce-driven medical system, uneven access to specialists, the reluctance people often have about approaching a doctor, all that now gets fed by a surrogate which should come with its own statutory warning. It could be erroneously bringing alive the “worst-case scenario” in your head. Online advisory trails can vary highly in quality and credibility. It can breed a cavalier attitude towards Ozempic or steroid-based creams. A pop-up quiz on Facebook can persuade you that you have ADHD, leading you to dubious apps, supplements or over-the-counter pills.
The answer is not to declare technology the villain and log off. Used correctly, it can be of much value. Our cover story this week, by Senior Editor Sonali Acharjee, equips you with the right habits: treat your next AI diagnosis as a very provisional prompt, not a final verdict. Do some hard counter-questioning. It’s a tool meant to help you. Not to keep you in a state of perpetual panic.
