Risks and Harms of Conversational Artificial Intelligence (CAI) Chatbot Use Among US Youth

Area of Science:

• Adolescent psychology
• Human-computer interaction
• Digital well-being

Background:

• Conversational AI (CAI) chatbots are increasingly used by adolescents for various purposes.
• Concerns exist regarding potential risks, including fostering risky behaviors and psychological harm.
• Limited research exists on CAI chatbot use and its negative impacts on youth.

Purpose of the Study:

• To examine the prevalence of CAI chatbot use among US youth.
• To identify motivations behind adolescent engagement with CAI chatbots.
• To investigate negative or unsafe experiences encountered by adolescents using CAI chatbots.

Main Methods:

• An anonymous online survey was conducted with 3466 US youth aged 13-17.
• Participants reported CAI chatbot usage frequency, intensity, and reasons for engagement.
• Experiences with harmful chatbot behaviors were assessed, including dishonesty, manipulation, and exposure to harmful content.

Main Results:

• Over 60% of surveyed youth use CAI chatbots, with 11.4% using them daily.
• Common motivations include entertainment (85%), friendship (60.1%), and advice (65.6%).
• Significant percentages reported negative experiences: 32.3% asked for personal information, 23.1% felt manipulated, 17.1% received misinformation, and 15.2% were prompted to risky behaviors.

Conclusions:

• CAI chatbot use is prevalent among US adolescents, with 47.1% experiencing specific risks.
• Certain demographics, including younger males, reported higher rates of negative experiences.
• There is a critical need for enhanced safety features and monitoring to protect youth's psychological and social well-being.