The bill would require companies to implement disclosures, design limits, and data protections for AI systems used by minors.
Senator Edward J. Markey (D-Mass.) introduced the Youth AI Privacy Act (Senate Bill S.4199), proposing specific requirements for companies that make AI systems available to minors, citing increased use of AI by children and the absence of consistent safety protections.
The bill applies to companies that develop, deploy, or operate AI systems accessible to users under 18 in the United States. The bill would require covered companies to:
- Disclose that users are interacting with an AI system, not a person
- Warn minors about the potential risks of using AI tools, including emotional reliance or harmful interactions
- Limit features that encourage prolonged or dependent use (e.g., push alerts) that could create emotional reliance
- Not train AI models on data from minors
- Ban AI tools from profiling a minor with their data
- Restrict the use of data collected via AI chatbots to direct user outputs or address safety issues within the chatbot
- Only use recently collected data when responding to a minor
- Not show ads to minors
- Implement safeguards to prevent harmful or inappropriate interactions, including self-harm, eating disorders, and substance use.
Senator Markey released a one-sheet accompanying the bill, “[Given] the explosive growth in AI chatbot usage, Congress must pass guardrails to protect our young people. The Youth AI Privacy Act does that by establishing new requirements on AI chatbot design features and requiring AI companies to implement critical privacy safeguards.”
The bill has been introduced in the Senate and referred to the Committee on Commerce, Science, and Transportation. It has not yet advanced to a vote.
