Chatbot Substance Advice Is Reaching Young Adults

Chatbot Substance Advice is quietly becoming a first stop for young adults with questions about drugs and alcohol, and researchers warn that the confidence of those answers may be part of the risk. A Rutgers study found that more than 32% of surveyed young adults had used a chatbot for information about substances, most often to learn about risks and consequences. The findings, reported by Rutgers University, land at a moment when chatbots are already woven into homework, job searches, and everyday health questions, and when a wrong answer about a substance can carry consequences that a wrong movie recommendation never will.

The study drew on survey data from nearly 1,000 participants aged 18 to 29 and was published in the Journal of Medical Internet Research Formative Research. About half of those who had used a chatbot for this purpose said they did so at least monthly, which suggests the behavior is not a one-off curiosity but a recurring habit for many users. Researchers also found that Black and male participants were more likely to turn to chatbots for substance-related questions, a pattern the lead author linked to possible differences in healthcare access, stigma, trust, and how people seek health information.

Why young adults turn to chatbots first

Part of the appeal is obvious: a chatbot answers instantly, privately, and without an appointment, a waiting room, or a judgmental look. For a young adult who is unsure whether a question is embarrassing, urgent, or even worth asking a clinician, that privacy can feel like safety. Broader survey data cited by the researchers show that 66% of young adults already use artificial intelligence to search for information and to get medical advice, so substance questions are arriving through a channel that many readers already trust for other topics.

That trust is itself a finding. Respondents in the study reported high levels of trust, reliability, and helpfulness for chatbot tools, according to the researchers. The problem is that fluency is not accuracy. A chatbot can present an invented dosage range, downplay withdrawal danger, or soften a warning in response to a leading question, all in complete, reassuring sentences with no visible uncertainty. For Chatbot Substance Advice, the danger is not only that an answer might be wrong, but that it might sound so polished that a reader stops checking it against a clinician, a pharmacist, or a trusted harm-reduction resource.

Where confident answers do the most damage

The most common reason participants gave for asking was unglamorous and understandable: they wanted to know about risks and consequences. That is also the category where a confident wrong answer does the most damage, because a reader is often deciding in the moment whether to mix substances, how much to take, or whether a symptom needs urgent care. A general-purpose assistant has no patient history, no way to examine the person asking, and no duty of care, yet it answers as if it understands the full picture. The gap between that confident tone and that limited knowledge is where harm can slip through.

The study's lead author, Mireia Triguero Roura, a postdoctoral researcher at Rutgers, 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 warning is not that every chatbot answer is wrong, but that the stakes are highest for the very group most likely to rely on them. A young adult who receives a reassuring answer about mixing alcohol with another drug may not seek a second opinion, and there may be no second opinion built into the chat window to prompt one.

There is also a counterpoint worth holding alongside the warning. For some readers, a chatbot may be the only source they will actually consult, especially if cost, stigma, or a lack of nearby services keeps them away from clinics. A careful answer that encourages professional help, names warning signs, and points toward local services could do real good for someone who would otherwise search alone. The researchers' argument is therefore not simply to ban the tool, but to understand how young adults interpret its outputs and to build better guardrails around the highest-risk questions.

What safer use looks like

Readers who do use a chatbot should treat its answer as a starting point, not a verdict. Check any dosage, interaction, or withdrawal claim against a clinician, a pharmacist, a poison control resource, or an established harm-reduction organization, and treat any answer that minimizes risk as a reason to check harder, not to relax. If a question involves a possible overdose, severe withdrawal, or a dangerous combination, skip the chat window entirely and seek urgent help. Those habits do not require abandoning a useful tool; they require refusing to let a polished paragraph stand in for medical judgment.

For Chatbot Substance Advice, the wider lesson reaches beyond one study. Artificial intelligence is rapidly becoming a layer between young adults and health information, joining search engines, friends, and school health classes, and it is doing so faster than public guidance has kept up. Developers, universities, and health services all have a role in that gap: clearer warnings on high-risk prompts, visible encouragement to consult clinicians, and campus resources that are as easy to reach as a chat window. Until those guardrails exist, the safest stance is a skeptical one, a point underscored by other recent reporting on how regulated organizations are handling artificial intelligence, new fronts in brain health research, and the uncertain job market young adults are navigating.

The researchers suggest further work should explore how young adults interpret and understand chatbot outputs about drugs and alcohol, and what those interpretations mean for public health. That research agenda matters because the technology will not wait for it: each month, more young adults will ask a chatbot a question they once asked a friend, a teacher, or a search engine. Whether those answers help or hurt will depend less on whether chatbots exist, and more on whether readers, developers, and health professionals treat a confident answer about substances as something to verify, every single time.