A complete substitute amendment. v2 significantly narrows the scope of the bill's regulations and prohibitions to focus primarily on "AI companions" made publicly available to end consumers, rather than all "artificial intelligence chatbots" as in v1. Penalties for criminal offenses and civil violations are increased, and new restrictions on third-party age verification providers are introduced.
GUARD Act
- Sponsor
- Sen. Hawley, Josh [R-MO]
- Committees
- Judiciary Committee (primary)
- Last action
- May 11, 2026
Bottom line
This bill aims to protect minors from harmful AI chatbot interactions by imposing strict age verification, content restrictions, and disclosure requirements on AI companion providers and all AI chatbots.
What it actually does
The GUARD Act of 2026 mandates age verification for AI companions, which are AI chatbots designed to simulate sustained interpersonal or emotional relationships, and prohibits minors from accessing them. It also requires all publicly available AI chatbots to disclose their non-human status and non-professional capacity. Furthermore, the bill establishes new federal criminal penalties for making AI chatbots publicly available that engage minors in sexually explicit conduct or promote self-harm, violence, or homicide.
Proponents argue
Proponents argue the bill is crucial for protecting children from the unique risks posed by advanced AI chatbots, including exposure to harmful content, emotional manipulation, and exploitation. They emphasize the compelling governmental interest in safeguarding minors from psychological and physical safety risks, asserting that current safeguards are insufficient to address the evolving capabilities of AI.
Opponents contend
Opponents might argue that the age verification requirements are overly burdensome, could infringe on privacy rights by requiring sensitive personal data, and may not be fully effective in preventing minor access while potentially limiting legitimate use cases for AI. They might also raise concerns about the scope of 'AI companion' and 'artificial intelligence chatbot' definitions, potentially stifling innovation and creating significant compliance costs for businesses.
The bill is relatively concise and well-structured, allowing for a thorough understanding within a reasonable timeframe for an informed reader.
Section 1
Short Title
Establishes the official short title of the Act as the 'Guidelines for User Age-verification and Responsible Dialogue Act of 2026' or the 'GUARD Act.' This provides a common, easy-to-reference name for the legislation, facilitating public discourse and legislative tracking.
Supporters argue
Supporters argue that a clear and concise short title helps in public discourse and legislative tracking, making the bill more accessible and understandable to a wider audience.
Critics contend
No specific opposition is typically raised against short title provisions as they are administrative and do not alter substantive policy.
Tradeoffs
None.
Section 2
Congressional Findings on AI Companion Risks to Minors
Outlines Congress's findings that AI companions are increasingly used by minors, can disseminate harmful or sexually explicit content, manipulate emotions, and pose risks like grooming, addiction, self-harm, and harm to others. It asserts that protecting children from such AI is a compelling governmental interest, laying the groundwork for the bill's regulatory framework.
Section 3
Key Definitions for AI Chatbots and Age Verification
Defines critical terms such as 'AI companion' (an AI chatbot simulating sustained interpersonal/emotional relationships or presenting as a sentient being), 'artificial intelligence chatbot' (interactive service generating non-predetermined, adaptive responses), 'covered entity' (person making AI chatbots publicly available), 'minor' (under 18), 'reasonable age verification measure' (e.g., verified ID, financial account age info), and 'reasonable age verification process' (using such measures, not just self-attestation). These definitions are crucial for determining the scope of the bill's obligations.
Section 4
Criminal Offenses for Harmful AI Chatbot Content to Minors
Establishes new federal crimes under 18 U.S.C. Chapter 6. It makes it unlawful to make an AI chatbot publicly available, knowing or with reckless disregard, if it engages minors in, erotically describes, or simulates sexually explicit conduct or obscene material, or makes such material available to minors. It also criminalizes making an AI chatbot publicly available that solicits, induces, or coerces minors to commit suicide, self-injury, unlawful physical/sexual violence, or homicide. Penalties are up to $250,000 per offense, providing a strong deterrent against severe harms.
Section 5(a)
Mandatory User Accounts for AI Companions
Requires covered entities to mandate user accounts for individuals accessing AI companions. This provision serves as a foundational step for implementing subsequent age verification requirements, ensuring that all users of these specific AI systems can be identified and their age verified.
Section 5(b)
Comprehensive Age Verification Process for AI Companions
Mandates a 'reasonable age verification process' for all AI companion user accounts. This includes freezing existing accounts until age is verified, verifying age for new accounts, and periodic re-verification. Covered entities can use third parties but remain liable, with a specific prohibition on contracting with third parties from 'covered nations' (e.g., China, Russia). It also imposes strict data security, retention limits, and a prohibition on sharing or selling age verification data, aiming to protect user privacy while ensuring age gating.
Section 5(c)
Disclosure Requirements for All AI Chatbots
Requires all publicly available AI chatbots to clearly and conspicuously disclose their non-human status at the start of each conversation and be programmed not to claim to be human. It also prohibits chatbots from representing themselves as licensed professionals (e.g., therapist, lawyer) and requires disclosures that they do not provide professional services, advising users to consult licensed professionals for such advice. This aims to prevent deception and manage user expectations.
Section 6
Prohibition of Minor Access to AI Companions
Explicitly prohibits minors from accessing or using any AI companion made publicly available by a covered entity, once their age has been determined through the verification process in Section 5(b). This is a core protective measure, directly preventing underage individuals from interacting with AI systems identified as posing specific risks.
Section 7
Enforcement Mechanisms and Penalties
Grants the Attorney General authority to bring civil actions to enjoin violations of Sections 5 or 6, enforce compliance, and obtain civil penalties (up to $250,000 per violation) and restitution. It also authorizes the Attorney General to promulgate necessary regulations, with specific limitations to prevent overreach. States are also empowered to bring civil actions as parens patriae on behalf of their residents. The Act clarifies it does not preempt more protective state laws, allowing for a layered enforcement approach.
The definition of 'AI companion' is crucial, specifying an AI chatbot that simulates a sustained interpersonal relationship or emotional interaction with the user, or presents at least one persistent identity and holds itself out as a sentient being, fictional character, or social entity.
Section 3(1)
Why it matters:It is standard legislative practice to define terms upfront, but the nuance here is critical as it significantly narrows the scope of the bill's most stringent requirements (age verification, minor prohibition) to a specific type of AI chatbot, rather than all AI chatbots.
Case for: Supporters argue this precise definition ensures the bill targets only the AI systems most likely to pose risks of emotional manipulation and inappropriate interaction with minors, avoiding overregulation of general-purpose chatbots and focusing resources where they are most needed.
Case against: Critics might argue the definition is still subjective and could lead to inconsistent application or capture AI not intended for regulation, potentially stifling innovation in areas like educational or therapeutic AI that might involve sophisticated interaction.
Estimated impact: Significantly limits the number of AI chatbots subject to age verification and minor prohibition, focusing the bill's most impactful provisions on a narrower, more targeted set of services.
Prohibits covered entities from contracting with a third party for age verification measures if that third party is closely affiliated with, incorporated in, organized under the laws of, or with a principal place of business in, a 'covered nation' as defined in section 4872 of title 10, United States Code.
Section 5(b)(4)(B)
Why it matters:While it is a technical detail of compliance, this provision has significant geopolitical and national security implications. Its placement might be considered standard legislative drafting for technical constraints, but it could easily be overlooked by a casual reader focused on the main age verification mandates.
Case for: Supporters would argue this provision is a critical national security safeguard, protecting sensitive age verification data from potential exploitation or access by foreign adversaries, particularly those identified as posing risks to U.S. interests.
Case against: Critics might argue this limits competition among age verification providers, potentially increasing costs or reducing options for covered entities, especially smaller businesses. It could also be seen as a protectionist measure or one that complicates global operations for AI companies.
Estimated impact: Restricts the market for age verification services for U.S.-facing AI companies, potentially impacting companies with ties to nations like China, Russia, Iran, and North Korea (as per 10 U.S.C. 4872).
Limits the Attorney General's rulemaking authority, stating that no regulation promulgated under the Act may impose a requirement not expressly authorized by the Act or address the development, deployment, or general operation of artificial intelligence systems outside the scope of obligations and offenses set forth in the Act.
Section 7(b)(2)
Why it matters:It's a common legislative mechanism to prevent executive agencies from exceeding the specific intent of Congress. While standard, its placement means a casual reader might miss the extent to which the AG's future regulatory actions are constrained.
Case for: Supporters would argue this provision prevents executive overreach and ensures that the Attorney General's regulations remain strictly within the bounds set by Congress, protecting against unforeseen regulatory burdens and maintaining legislative control over AI policy.
Case against: Critics might argue this limits the flexibility of regulators to adapt to rapidly evolving AI technologies and unforeseen challenges that may emerge, potentially making the law less effective over time without further legislative action to update it.
Estimated impact: Constrains the future regulatory scope of the Attorney General regarding AI, ensuring a more static implementation of the Act and requiring new legislation for significant policy shifts.