Understanding Generative AI Risks for Youth: A Taxonomy Based on Empirical Data

2025-02-22

AI Summary

Generative AI (GAI) poses emerging risks to adolescents’ mental health, behavioral development, and social adaptation—particularly novel forms of toxicity, privacy violations, and exploitation that remain unaddressed by current child online safety and AI governance frameworks. Method: We developed the first empirically grounded taxonomy of adolescent GAI risks through a mixed-methods approach—qualitative coding, thematic modeling, cross-source triangulation, and interaction-path mapping—based on 344 real-world GAI dialogues, 30,300 Reddit discussions, and 153 documented AI-related incidents. Contribution/Results: The taxonomy identifies 84 distinct risk types organized across six dimensions and mapped onto four primary human–AI interaction pathways. It bridges a critical gap in the intersection of adolescent digital well-being and AI risk research, providing a structured evidence base to inform protective mechanism design by developers, digital literacy interventions by educators, and updated AI safeguards for minors by policymakers.

Technology Category

• Philosophy and Ethics of AI: Safety, Robustness & Trustworthiness
• Humans and AI: Game Design — Procedural Content Generation & Storytelling
• Application Domains: Humanities & Computational Social Science

Application Category

• Social Networks and Social Media: Generative AI / large language models and their impact on social systems
• Economics, Online Markets and Human Computation: Economic ramifications for generative AI infrastructure and applications
• Semantics and Knowledge: Data modeling to support human-machine intelligence, including LLMs agents, intelligent system behavior, explanations, and user-friendly interactions

Abstract

Generative AI (GAI) is reshaping the way young users engage with technology. This study introduces a taxonomy of risks associated with youth-GAI interactions, derived from an analysis of 344 chat transcripts between youth and GAI chatbots, 30,305 Reddit discussions concerning youth engagement with these systems, and 153 documented AI-related incidents. We categorize risks into six overarching themes, identifying 84 specific risks, which we further align with four distinct interaction pathways. Our findings highlight emerging concerns, such as risks to mental wellbeing, behavioral and social development, and novel forms of toxicity, privacy breaches, and misuse/exploitation that are not fully addressed in existing frameworks on child online safety or AI risks. By systematically grounding our taxonomy in empirical data, this work offers a structured approach to aiding AI developers, educators, caregivers, and policymakers in comprehending and mitigating risks associated with youth-GAI interactions.

Problem

• Identifies risks in youth-GAI interactions
• Categorizes 84 specific risks empirically
• Addresses mental wellbeing and privacy breaches

Innovation

• Empirical data taxonomy
• Youth-GAI risk categorization
• Systematic risk mitigation