Rutgers study finds one in three surveyed young adults used AI for substance-use information

Young adults seeking information about substance-use risks are turning to chatbots they describe as trustworthy and helpful. In a study announced by Rutgers on October 23, more than 32% of nearly 1,000 U.S. participants ages 18–29 reported using ChatGPT or similar tools for substance-use information—raising researchers’ concerns about reliance on answers that can contain misinformation.

The study, published in JMIR Formative Research, examined who uses generative AI for substance-related questions, how often they turn to it and what they think of the results. Generative AI includes chatbots that produce text in response to a user’s question or instruction.

Participants primarily wanted information about risks and consequences. That places the reported use in a safety-relevant setting: respondents were asking about possible harms, rather than merely describing general interest in AI. The finding identifies what they sought, not whether the answers they received were sound.

For many of those users, this was not a one-time inquiry. About half of respondents who used AI for substance-use information said they did so at least monthly. That frequency applies to the subgroup already using chatbots for this purpose, not to half of everyone surveyed.

Trust is a perception, not an accuracy score

Researchers found that young adults in the study perceived AI tools as highly trustworthy, reliable and helpful.

The cross-sectional survey captures reported use and attitudes at one point, rather than tracking how chatbot advice changes behavior over time. Its clearest result is therefore about the audience already seeking this information and the confidence they place in AI—not a measured health benefit or harm.

Different users, unresolved reasons

Black participants and male participants were more likely to report using AI for substance-use purposes.

Triguero Roura suggested several possible explanations: differences in health care access, stigma, trust and the ways people seek health information.

Triguero Roura and co-author Pia M. Mauro called for further research into how young adults interpret and understand AI outputs about drugs and alcohol, and what that means for public health.