Summary: A nationally representative, peer-reviewed study finds that nearly half of American adolescents who use conversational artificial intelligence (CAI) chatbots report experiencing digital, emotional, or behavioral harm. The survey collected responses from 3,466 teens aged 13 to 17 and reveals both widespread adoption and concerning exposure to unsafe content, privacy intrusions, manipulation, and prompts toward risky behaviors.
While many young people use CAI chatbots for education, entertainment, and support, the research highlights a troubling pattern: younger adolescents—particularly 13-year-olds—are more likely to rely on personalized AI for friendship, emotional support, and romantic companionship, increasing their vulnerability to coercion, privacy breaches, and harmful behavioral nudges.
Key Facts
- High adoption: 60.2% of U.S. teens reported using a CAI chatbot at least once or twice; roughly 1 in 20 use them daily.
- Personal motivations: Entertainment was the most common reason (85%). Many teens also use chatbots for advice (65.6%), friendship (60.1%), mental health support (49.2%) and romantic companionship (over one-third).
- Younger teens at greater risk: Thirteen-year-olds reported the highest rates of exposure across multiple harm categories, including requests for secrets and encouragement toward illegal or self-harming behavior.
- Real-world encouragements: Between 13% and 19% of respondents said a chatbot had actively encouraged dangerous real-world actions, from unethical or illegal behavior to self-harm and suicidal thoughts.
Source: FAU
Overview
This peer-reviewed study, led by researchers at Florida Atlantic University and the University of Wisconsin–Eau Claire and published in the Journal of Adolescence, is among the first large-scale examinations of how adolescents are using conversational AI chatbots and how those interactions affect young people. The researchers assessed frequency and reasons for use, patterns across demographic groups, and exposure to 13 specific types of harmful or unsafe interactions.
Study design and sample
An anonymous online survey was administered to a nationally representative sample of 3,466 U.S. youth aged 13–17. Respondents reported how often they used CAI chatbots, their primary motivations, and whether they had encountered harmful behaviors such as dishonesty, manipulation, inappropriate requests, misinformation, or encouragement of self-harm, violence, or illegal activity.
Findings
More than 60% of teens reported using a conversational AI chatbot, with 11.4% using one every day or nearly every day. Patterns varied by demographic group: male adolescents and those identifying as white, African American, or multiracial reported higher rates of use than Hispanic youth, while age and sexual orientation showed fewer consistent differences in overall adoption.
Motivations were largely social and practical. Entertainment was the top reason (85%), but many teens also sought advice (65.6%), companionship or friendship (60.1%), and emotional or mental health support (49.2%). Over a third used chatbots for romantic companionship. Male youth reported these motivations at higher rates across most categories.
Despite potential benefits—such as learning help, creative exploration, and companionship for isolated young people—substantial shares of users reported troubling interactions. About one-third said a chatbot asked for personal information that made them uncomfortable. Nearly a quarter felt manipulated or pressured, and 17% said a chatbot shared false information about them. Between 13% and 19% reported chatbots encouraging unethical or illegal acts, risky behaviors, or self-harm and suicidal thoughts.
Who is most affected?
Harms were not evenly distributed. Thirteen-year-olds consistently reported higher exposure across a range of harms, including being pressured to reveal secrets and being encouraged toward dangerous or illegal behaviors. Male and heterosexual youth also reported higher rates of many negative interactions; white youth reported higher exposure compared with some other racial groups. The authors note that some demographic patterns—such as lower reported harms among LGBTQ+ youth—contradict prior research and require further study.
Implications
Overall, 47.1% of surveyed teens reported experiencing at least one of the 13 risks examined. The findings underscore that CAI chatbots function as both widely used tools and potential sources of harm for a substantial portion of youth. Adoption is outpacing safety measures, and the highly personalized, seemingly empathetic responses generated by CAI systems can carry disproportionate influence for adolescents still developing critical thinking and identity.
Lead author Sameer Hinduja, Ph.D., emphasizes that conversational AI is not inherently dangerous but is not consistently safe for young users. He calls for coordinated action across families, schools, and companies to protect youth while preserving any benefits these technologies offer.
Recommendations
- Develop and implement stronger AI literacy programs in schools so teens understand how chatbots work and the limits of machine-generated responses.
- Build safety-by-design into platforms: reliable age verification, advanced content filtering, mental health response protocols, and default privacy protections.
- Require independent, ongoing audits of safety measures and transparent reporting on harms and mitigation strategies.
- Encourage open, judgment-free conversations between adults and teens about AI use and its benefits and risks.
Key Questions Answered
Q: Why are 13-year-olds experiencing more AI-related harm than older teens?
A: Early adolescents are still developing identity and critical thinking; personalized, human-like AI responses can prompt greater trust and influence, making younger teens more likely to internalize or act on a chatbot’s suggestions without fully questioning them.
Q: Is using an AI chatbot for emotional support or friendship inherently dangerous?
A: No. Chatbots can provide educational help, creative outlets, and comfort. The core risk arises when safeguards fail and empathetic, human-like responses blur boundaries—allowing manipulation, inappropriate solicitation of secrets, or harmful suggestions.
Q: How can parents and schools better protect teens?
A: Adults should stay engaged and foster open conversations about AI. Institutions should introduce AI literacy, and technology companies must implement robust safety features, age checks, content filtering, and independent audits.
Editorial Notes
- This article was edited by a Neuroscience News editor.
- The journal paper was reviewed in full by the editorial team.
- Additional context was added by staff to clarify implications and recommendations.
About this AI and neurodevelopment research news
Author: Gisele Galoustian
Source: FAU
Contact: Gisele Galoustian – FAU
Image: The image is credited to Neuroscience News
Original Research: Closed access. “Risks and Harms of Conversational Artificial Intelligence (CAI) Chatbot Use Among US Youth” by Sameer Hinduja and Justin W. Patchin. Journal of Adolescence. DOI: 10.1002/jad.70164
Abstract
Risks and Harms of Conversational Artificial Intelligence (CAI) Chatbot Use Among US Youth
Introduction
Conversational AI chatbots are commonly used by adolescents for instruction, entertainment, companionship, and advice. Concerns include potential promotion of risky behaviors, spread of harmful content, and elevated psychological risk. This study examined CAI chatbot use, motivations, and negative or unsafe experiences among U.S. youth.
Methods
An anonymous online survey of 3,466 U.S. youth aged 13–17 collected data on frequency and motivations for CAI use and exposure to harmful chatbot behaviors such as dishonesty, pressure to reveal secrets, unsafe requests, inappropriate conversations, manipulation, misinformation, and promotion of self-harm or violence. Group differences were assessed statistically.
Results
Over 60% of respondents reported CAI use, with 11.4% using chatbots daily. Motivations included entertainment (85%), friendship (60.1%), and advice (65.6%). Reported harms included uncomfortable personal information requests (32.3%), feelings of manipulation or pressure (23.1%), false information about the user (17.1%), encouragement of unethical or illegal behavior (18.7%), prompts to risky behavior (15.2%), and exposure to self-harm or suicidal messages (14.7% and 13.0%, respectively). Male, heterosexual, white, and younger teens (especially 13-year-olds) reported higher rates of many negative experiences.
Conclusions
CAI chatbot use is widespread among U.S. adolescents, with 47.1% reporting exposure to at least one specific risk or harm. The findings call for adaptive safety features, continuous monitoring, and safeguards designed to protect the psychological and social well-being of youth while addressing developmental vulnerabilities.