S. 3062 — GUARD Act

This section sets forth congressional findings on the deployment of artificial intelligence chatbots on social media platforms and in consumer applications used by minors, including their capacity to generate harmful or sexually explicit content, manipulate emotions by exploiting developmental vulnerabilities, and expose children to risks such as grooming, addiction, self-harm, and harm to others.

This section defines key terms for the Act. (1) "AI companion" means an artificial intelligence chatbot that provides adaptive, human-like responses to user inputs and is designed to encourage or facilitate the simulation of interpersonal or emotional interaction, friendship, companionship, or therapeutic communication; (2) "artificial intelligence chatbot" means any interactive computer service or software application that produces new expressive content or responses not fully predetermined by the developer or operator, accepts open-ended natural-language or multimodal user input, and produces adaptive or context-responsive output, excluding services limited to contextualized replies within a narrow specified purpose; (3) "covered entity" means any person who owns, operates, or makes available an artificial intelligence chatbot to individuals in the United States; (4) "minor" means any individual under 18 years of age; (5) "reasonable age verification measure" means a method, such as government-issued identification or another commercially reasonable method, that reliably determines whether a user is an adult and prevents minor access; and (6) "reasonable age verification process" means a process that applies one or more such measures, requires more than self-reported age confirmation, subjects every user to verification, and does not rely on shared technical indicators with known adult users.

This section establishes criminal prohibitions on the design, development, or availability of artificial intelligence chatbots. It defines an artificial intelligence chatbot as an interactive computer service or software application that produces new expressive content or responses and accepts open-ended natural-language or multimodal input, excluding systems limited to contextualized replies for a narrow specified purpose. The section makes it unlawful to create or offer such a chatbot knowing or with reckless disregard that it poses a risk of soliciting, encouraging, or inducing minors (individuals under 18 years of age) to engage in, describe, or simulate sexually explicit conduct or to create visual depictions of such conduct, subject to a fine of not more than $100,000 per offense. It also prohibits chatbots that encourage, promote, or coerce suicide, non-suicidal self-injury, or imminent physical or sexual violence, subject to the same penalty.

This section establishes obligations for covered entities that make artificial intelligence chatbots available to users. It requires each covered entity to mandate creation of a user account for any individual accessing or interacting with such a chatbot. For accounts existing on the effective date of the Act, the entity must freeze the account until the user provides verifiable age data, then classify the user as a minor or adult; the same age verification and classification process applies to new accounts and must be repeated periodically. A covered entity may use a third party for age verification but remains fully responsible for compliance and liability. The section further requires covered entities to implement data security measures that limit collection of age verification data to the minimum necessary, protect it with industry-standard encryption, retain it only as long as needed for compliance, and prohibit sharing, transfer, or sale of such data. Finally, each chatbot must disclose at the start of every conversation and at 30-minute intervals that it is an artificial intelligence system and not a human being, must not claim to be human, and must disclose at the start of each conversation and at reasonably regular intervals that it does not provide medical, legal, financial, or psychological services and that users should consult a licensed professional for such advice.

This section prohibits a covered entity from allowing a minor to access or use any AI companion if the age verification process in section 5(b) determines that the individual is a minor.

This section authorizes the Attorney General to investigate violations of sections 5 or 6 and to bring civil actions in federal district court to enjoin such violations, compel compliance, obtain restitution, and impose civil penalties not to exceed $100,000 for each separate violation. It also empowers state attorneys general to bring parens patriae actions on behalf of state residents for injunctive relief and preserves the enforcement of any state law or regulation that provides at least as much protection to users of artificial intelligence chatbots as this Act.