The "Guidelines for User Age-verification and Responsible Dialogue Act of 2025" aims to regulate artificial intelligence chatbots that interact with minors. It addresses concerns about chatbots disseminating harmful or sexually explicit content, manipulating emotions, and influencing vulnerable minors. The Act defines terms like AI companion, artificial intelligence chatbot, covered entity, and minor. It mandates reasonable age verification measures, prohibits chatbots from claiming to be human or professionals, and enforces penalties for violations. The Act also imposes obligations on covered entities regarding user accounts, age verification, and disclosures, while prohibiting minors from using AI companions. Enforcement involves civil actions and penalties. The Act will take effect 180 days after enactment, aiming to protect minors from the potential risks associated with these chatbots.
Guidelines for User Age-verification and Responsible Dialogue Act of 2026 or the GUARD Act
This bill establishes requirements for and limitations on entities that provide publicly available artificial intelligence (AI) chatbots, including AI companions.
Under the bill, an AI chatbot is an application or program that simulates human conversation by generating responses to user prompts. An AI companion is a type of AI chatbot that simulates a sustained interpersonal relationship or emotional interaction with the user.
The bill makes it a federal crime to provide an AI chatbot that exposes minors to sexual content or encourages minors to commit suicide or carry out nonsuicidal self-injury, unlawful physical or sexual violence, or homicide. A violation is subject to a criminal fine of up to $250,000 per offense.
Additionally, the bill prohibits minors from accessing AI companions. Under the bill, an entity that provides an AI chatbot must require each individual who accesses an AI companion to set up a user account and must also verify that the user is an adult. A violation is subject to a civil penalty of up to $250,000 per violation.
Finally, the bill requires an AI chatbot that is publicly available to disclose that the AI chatbot is not a human and is not a licensed professional (e.g. doctor, lawyer, or therapist). A violation is subject to a civil penalty of up to $250,000 per violation.