Bill > HB1261

This bill, the Artificial Intelligence Toy Safety Act, establishes new requirements for toys that use artificial intelligence (AI) to ensure child safety and protect their data privacy. It defines an "artificial intelligence toy" as any product sold to consumers in the state that uses AI or similar computational processes and is marketed to or primarily used by a child. The bill prohibits these toys from generating inappropriate content, such as violent or sexually explicit material, or content that is emotionally manipulative, and also prevents them from encouraging children to interact with third-party online systems. Manufacturers must implement mechanisms for parental consent, disable AI features if consent is not obtained, include age-appropriate filters to block harmful dialogue, and have content moderation to prevent prohibited outputs. Additionally, AI toys must have an automatic "safe mode" that activates when potentially harmful input is received and be capable of secure software updates. Manufacturers are required to conduct a child AI safety assessment before selling a toy, annually thereafter, and after any significant AI feature changes, outlining foreseeable risks and mitigation strategies. The bill also restricts manufacturers to collecting only the minimum "child user data"—information collected or inferred from a child's interaction with the toy—necessary for the toy's core functions, mandates encryption of this data, and prohibits its sale, lease, or transfer to third parties, use for training unrelated AI models, or use for targeted advertising. Child user data can only be retained for 12 months unless parental consent is renewed, and parents must be able to easily access, review, download, and delete their child's data, as well as disable data collection without affecting the toy's core functionality. In the event of a data breach, manufacturers must notify affected parents or guardians within 48 hours. AI toys must also carry a clear label indicating they use AI and providing information about data collection and how interactions are stored or used, and manufacturers must make an easily accessible AI safety summary report publicly available. Furthermore, manufacturers cannot market AI toys as emotional companions or parental substitutes, use recordings of child interactions in marketing without consent, or use collected data to develop advertisements targeted at children. A Child Artificial Intelligence Toy Safety Review Panel will be established within the Attorney General's office to review compliance, conduct audits, and make recommendations for safety standards and legislative changes. Violations of this subtitle are considered unfair, abusive, or deceptive trade practices and are subject to penalties under existing law, including civil penalties of up to $50,000 per violation and mandatory recalls. The act is set to take effect on July 1, 2026.

Business and Industry

https://mgaleg.maryland.gov/mgawebsite/Legislation/Details/HB1261?ys=2026RS