Addictive Design Act of 2026
- Sponsor
- Rep. Balint, Becca [D-VT-At Large]
- Committees
- Education and Workforce Committee (primary)
- Last action
- Jul 27, 2026
Bottom line
The bill aims to protect youth mental health by researching the impact of AI chatbots, educating adults, and directly regulating AI chatbot design features deemed addictive, with significant penalties for non-compliance.
What it actually does
This bill directs the President to establish an interagency task force to review the impact of AI chatbots on youth mental health and recommend ways to reduce addictive design features. It also authorizes the Secretary of Health and Human Services to establish grant programs for research into AI chatbot impacts and for outreach to parents and teachers. Furthermore, the bill mandates that AI chatbot providers grant researchers access to anonymized data and prohibits them from making AI chatbots with defined 'addictive design features' available to youth, backed by civil penalties.
Proponents argue
Proponents argue that the bill is a critical step to address the growing mental health crisis among youth, which is exacerbated by intentionally addictive design features in AI chatbots. They contend that the bill's combination of research, education, and direct regulation is necessary to mitigate the documented risks, including unhealthy emotional attachments and potential links to suicide, while balancing commercial freedoms with public safety interests.
Opponents contend
Critics might argue that the bill's definition of 'addictive design features' is overly broad, potentially stifling innovation in AI development or capturing benign conversational elements. Concerns could also be raised about the feasibility and privacy implications of strict age verification requirements and the potential for the data access mandate to create undue burdens or security risks for AI chatbot providers. Some may also argue that the penalties are disproportionate or that the bill could face First Amendment challenges regarding commercial speech.
This bill is relatively short and straightforward, making it accessible for an informed reader to evaluate within a reasonable timeframe.
Section 3
Establishment of Interagency Task Force on AI Chatbots and Youth Mental Health
This section directs the President to establish an interagency task force within 60 days of the bill's enactment. The task force, composed of federal officials, academic researchers, and mental health professionals, will review existing research and make recommendations on minimizing mental health risks, identifying warning signs, and providing mental health services related to youth use of AI chatbots. The recommendations will be publicly available.
Supporters argue
Supporters argue that establishing an expert interagency task force is crucial for developing evidence-based policies to address the complex issue of AI's impact on youth mental health. They believe it ensures a comprehensive review and coordinated federal response.
Critics contend
Critics might suggest that establishing another task force could be a slow, bureaucratic process that delays urgent action, or that existing agencies already possess the mandate and expertise to conduct such reviews, potentially leading to redundant efforts.
Tradeoffs
The provision balances the need for thorough, expert review with the urgency of addressing youth mental health concerns. It also aims to bring together diverse perspectives, which can be challenging to coordinate effectively.
Section 4(a)
Research Grant Program on AI Chatbot Impact
This section directs the Secretary of Health and Human Services to establish a grant program for institutions of higher education. These grants will fund research projects specifically focused on the impact of AI chatbots on youth mental health, with the findings intended to inform the recommendations of the interagency task force.
Section 4(b)
Outreach Grant Program for Parents and Teachers
This section directs the Secretary of Health and Human Services to establish a grant program for eligible entities, such as non-profit organizations, to conduct outreach. This outreach will provide training and information to parents, legal guardians, and elementary, middle, and high school teachers on how to identify and respond to mental health risks, including unhealthy emotional attachment and addiction, associated with youth use of AI chatbots.
Section 5
Mandated Data Access for Research and Civil Penalties
This section requires covered entities (AI chatbot providers) to provide anonymized data and information to institutions of higher education that receive research grants under Section 4(a), upon request. Failure to comply with a request for a single research project is considered one violation, subject to a civil penalty of up to $10,000,000, and the Attorney General may seek injunctive relief.
Section 6
Prohibition on Addictive Design Features for Youth and Age Verification Requirements
This section prohibits covered entities from making AI chatbots with 'addictive design features' available to youth. It also mandates that if age verification technology is deployed to ensure compliance, it must limit data collection to only what is strictly necessary for age determination and delete that data within 24 hours. Violations of either the prohibition or the age verification requirements incur a civil penalty of $5,000 per event or transaction, with the Attorney General able to seek injunctive relief.
The bill broadly defines 'addictive design feature' to include elements such as showing typing bubbles, providing avatars, storing and using personal information from previous conversations, requiring additional engagement or pay to continue a conversation, maintaining a conversation for longer than 2 hours, or impersonating a real person, living or dead.
Section 7(1)
Why it matters:While standard legislative practice to define terms, the breadth and specificity of this definition are highly consequential for AI chatbot developers and could be easily missed in a quick read, despite its direct impact on the core prohibition.
Case for: Supporters argue that providing specific examples within the definition ensures clarity for enforcement and prevents developers from circumventing the intent of the law by making minor design changes. It aims to capture a wide range of manipulative design tactics.
Case against: Critics might argue that this definition is overly broad, potentially criminalizing common or benign features of conversational AI (e.g., typing bubbles, avatars) that are not inherently 'addictive.' They contend it could stifle innovation and make it difficult for developers to create engaging, helpful AI experiences for youth without risking penalties.
Estimated impact: This definition directly determines the scope of the prohibition on addictive design features, significantly impacting how AI chatbot providers must design and operate their services for youth users.
For purposes of assessing civil penalties for violating the prohibition on addictive design features or age verification requirements, 'failure to comply under subsection (a) or (b) with respect to the same event or transaction, or a related series of events or transactions, shall be considered as one violation.'
Section 6(c)(2)
Why it matters:This is technical language common in legislation to clarify enforcement mechanisms. However, for covered entities, understanding how violations are aggregated is crucial for assessing potential financial liability.
Case for: Supporters argue this clarifies the enforcement mechanism, ensuring that penalties are applied consistently and that companies cannot easily evade accountability by claiming each instance of non-compliance is a separate, minor event. It aims to deter systemic non-compliance.
Case against: Critics might argue that even with this aggregation, a $5,000 penalty per 'event or transaction' could still lead to astronomically high fines for large platforms with millions of users, potentially crippling companies for minor or unintentional infractions. They may also argue that 'related series of events or transactions' is vague and open to broad interpretation.
Estimated impact: This clause directly influences the potential financial liability for covered entities, determining whether a single design flaw affecting many users results in one or multiple violations, significantly impacting the scale of potential fines.