One in three young adults ask AI chatbots about drugs, alcohol: study

About one in three young adults in the United States have used AI chatbots such as ChatGPT, Google Gemini and Meta AI to look up information about drugs and alcohol, according to a Rutgers Health study reported on October 26, 2026. The researchers surveyed 995 adults aged 18 to 29 in April 2026 and found that 32% had used generative AI for substance-related questions. The paper was published on October 19 in the journal JMIR Formative Research. The lead author warned that relying on these tools could lead some young people into misinformed or risky choices.

Young people have always looked for answers about alcohol and drugs away from parents and doctors, first from friends, then from search engines and forums. This study suggests a new step: a large share of 18-to-29-year-olds now put those questions to a chatbot, and they trust the replies a little more than they trust AI for other topics. That matters because a chatbot gives a single, confident-sounding answer rather than a list of sources the reader can weigh. If the answer on a dose, a mix of substances or a withdrawal symptom is wrong, there is no second opinion on the screen.

The lead author, Mireia Triguero Roura, put the concern plainly: "Using these tools can lead to misinformed or otherwise risky substance-use behavior among young people, who are the age group most affected by adverse effects of substance use." The researchers did not say that chatbots give bad advice. They measured who uses them and why, and they raised the question of what those answers contain.

The study also has clear limits. It was a one-time online survey of Americans, so it shows what people reported at a single point rather than how their behaviour changed. It did not test whether the chatbots' answers were accurate, and it did not follow participants to see whether the advice changed what they did. The higher use among Black respondents and men is an association; the authors pointed to possible reasons such as access to health care, stigma, trust and search habits, but did not prove any of them.

This study did not survey India, but the same chatbots are open to Indian students and young workers, and the questions it raises apply here too. The practical point is the same anywhere: a chatbot can be a starting point for general facts, but questions about doses, mixing substances or quitting need a doctor, a pharmacist or a trained counsellor. In India, the Narcotics Control Bureau's MANAS helpline, 1933, is a 24-hour toll-free number for counselling and de-addiction support as well as for reporting drug trafficking.

Short-term: The findings add to calls for AI companies to make sure answers on drugs and alcohol are accurate and point users to professional help, and for health services to meet young people where they already search.

Long-term: Public health campaigns may need to treat chatbots as a main source of health information for young adults, alongside search engines and social media, and test what those tools actually say.

Who is affected: Students and young adults who use AI tools for health questions; parents, teachers and counsellors; doctors and addiction services; and companies that build chatbots.

Questions and Answers