What does Preamble of SAFE KIDS Act do?

This bill, formally titled 'To require providers of certain artificial intelligence systems to implement child safety by design, parental settings, and independent audits, to prohibit child targeted advertising and the sale or sharing of children’s personal information, and for other purposes,' aims to address child safety concerns related to artificial intelligence systems. The bill requires providers of certain AI systems to implement 'child safety by design' measures. It mandates the inclusion of parental settings within these AI systems. The legislation calls for independent audits of these AI systems. The bill seeks to prohibit child-targeted advertising. It also aims to prevent the sale or sharing of children's personal information.

What does SHORT TITLE. of SAFE KIDS Act do?

This bill is formally titled the Safeguarding AI Features to Ensure Kids' Informed Digital Safety Act.

What does DEFINITIONS. of SAFE KIDS Act do?

This section defines terms used throughout the Act. An 'AI chatbot' is defined as an artificial intelligence system that generates responses not fully predetermined and accepts open-ended user input, but excludes systems used solely for customer service, internal enterprise research, or limited contextual replies. A 'child safety risk' refers to any reasonably foreseeable risk of harm to a child. The term 'covered harm' includes physical harm, sexual exploitation, financial harm, psychological harm, privacy intrusions, and adverse discrimination proximately caused by AI chatbot use. A 'qualified researcher' is defined as someone affiliated with an academic or research institution who demonstrates expertise and a legitimate public-interest research purpose related to AI chatbot risks. “Provider of an AI chatbot” means a person who makes an AI chatbot available to a user in the United States.

What does DETERMINATION OF USER AGE. of SAFE KIDS Act do?

This section outlines requirements for AI chatbot providers regarding user age determination. Providers must generally treat users as children unless they have verified the user is 18 or older. Providers are required to implement age estimation technology to distinguish between child and adult accounts. Age signals from this technology are treated as the user's actual age range, with exceptions for conflicting information. Providers may use alternative age signals from operating systems or app stores, but must resolve conflicts by using the younger age. Providers must periodically review user accounts using age estimation technology. Providers must implement data security measures to protect age estimation data, limiting collection and prohibiting its use for purposes other than age estimation. Providers may be deemed compliant if they implement an age assurance framework comparable to existing foreign, federal, or state laws.

What does ADDITIONAL DUTIES OF A PROVIDER OF AN AI CHATBOT. of SAFE KIDS Act do?

This section outlines additional duties for providers of AI chatbots, particularly concerning child safety. Providers must conduct and document comprehensive risk assessments to identify child safety and privacy risks arising from their AI chatbots, at required intervals including before release to children, before significant updates, and annually. Risk assessments must consider the likelihood of harm, differential risks across age groups, known vulnerabilities of children, empirical data, and relevant research. Providers are required to implement measures to mitigate identified child safety risks and establish safeguards for child users, such as usage reminders, age-appropriate warnings, and protective design features. Providers are prohibited from making available AI chatbots that generate content promoting eating disorders, encouraging unlawful activities, or depicting harmful or exploitative content. Providers must publish a child safety policy on their website disclosing the risk assessment process, identified risks, implemented safeguards, and parental settings. A documented crisis-response protocol is required for conversations indicating a child is at risk of harm, including suicidal ideation, with timely referral to external resources and parental notification where appropriate. Providers must provide clear notice to child users that they are interacting with an AI system, reinforced periodically. Providers are restricted from creating AI chatbots that mislead children into believing they are interacting with a human or encourage inappropriate emotional dependence. Providers must offer parental settings allowing control over various aspects of the AI chatbot's interaction with a child, and actively promote these settings.

What does ADVERTISING AND INFORMATION SHARING PROTECTIONS FOR CHILDREN. of SAFE KIDS Act do?

This section prohibits AI chatbot providers from advertising to child users, including through product placement. It also prohibits child-targeted advertising by AI chatbot providers. The sale or sharing of a child's personal information by an AI chatbot provider is prohibited unless verifiable parental consent is obtained. An exception exists for disclosures necessary to respond to judicial process or provide information to law enforcement. AI chatbot providers cannot condition a child's use of the chatbot on parental consent for data sales or require excessive personal information. Providers are prohibited from using deceptive design patterns that mislead children or parents regarding safety features, privacy controls, or parental settings.

What does RULEMAKING. of SAFE KIDS Act do?

This section requires the Commission to establish an incident reporting mechanism for AI chatbot use by children within 180 days of the Act's enactment. The reporting mechanism will allow third parties, such as users, parents, educators, researchers, and advocacy organizations, to report incidents to the Commission. The Commission must publish standards defining reportable incidents, including content promoting harm, sexually explicit conduct, failures in age assurance, and deceptive outputs. The Commission is also required to establish and maintain a publicly accessible online resource linking to each AI chatbot's child safety policy within 180 days of enactment.

What does AI CHILD SAFETY AUDIT. of SAFE KIDS Act do?

This section establishes requirements for independent audits of AI chatbots to assess child safety risks. Each AI chatbot provider must submit an annual audit at their own cost, conducted by an auditor meeting specific professional and independence standards. Audits must test chatbot outputs using controlled accounts, evaluate safeguards, and assess compliance with the Act's requirements. Providers can demonstrate compliance by conducting audits aligned with comparable foreign, federal, or state laws. Providers must submit an AI child safety audit report to the Commission within 90 days of completion and publish a summary on their website, with redactions for sensitive information. The Commission will review audit reports, establish procedures for reviewing them, and publish aggregated findings annually to inform parents and policymakers. A public registry will be maintained with high-level summaries of audit reports. Qualified researchers may access anonymized audit data for academic study, subject to strict data use agreements and protections. The Commission may disclose audit information to government agencies for law enforcement purposes. The section clarifies that requirements do not compel disclosure of security vulnerabilities or decryption of user content.

What does ENFORCEMENT BY THE COMMISSION. of SAFE KIDS Act do?

This section outlines the enforcement mechanisms for the Act. Violations of this Act or its regulations are treated as violations of an unfair or deceptive act or practice under the Federal Trade Commission Act. The Commission is granted authority to enforce this Act similarly to the Federal Trade Commission Act. Individuals violating the Act are subject to penalties and privileges outlined in the Federal Trade Commission Act. Additional civil penalties apply for violations, including $1,000 per violation per user for failures to implement safeguards and $10,000 per violation per user for willful violations involving false information. The Commission retains its authority under other laws. The Commission can promulgate rules to carry out the Act, following Title 5, United States Code, section 553.

What does RELATIONSHIP TO OTHER LAWS; PREEMPTION. of SAFE KIDS Act do?

This section clarifies the relationship of this Act to other laws. The Act does not prohibit the enforcement of existing Federal or State laws that offer at least as much protection to users of AI chatbots as this Act. Compliance with this Act's requirements, including audits and risk assessments, does not exempt AI chatbot providers from liability for violating other laws. Compliance with this Act does not constitute a defense in civil or criminal actions under Federal or State law.

Back to bills

SAFE KIDS Act

Last updated: 6/23/2026 · Introduced: 6/23/2026

Author: John Curtis (R-UT)

TL;DR (AI)

The SAFE KIDS Act requires providers of AI chatbots to implement 'child safety by design' measures and include parental settings within their systems.
It prohibits AI chatbot providers from targeting advertising to children and from selling or sharing children’s personal information without verifiable parental consent.
Providers must conduct regular risk assessments to identify and mitigate potential harms to children resulting from AI chatbot use, including content promoting harmful behaviors.
The Act mandates independent annual audits of AI chatbots to assess child safety risks and ensure compliance with its requirements.
Violations of the Act can result in significant civil penalties, with the Federal Trade Commission responsible for enforcement.

119th Congressin committeeOfficial page

Verified Votes

No verified votes yet.

Community Votes

No community votes yet. Be the first!

Voting as guest · Sign in for verified votes

Other Sections

Includes provisions on preamble, short title., and 10 more.

12 sections

Click "Show details" to explore individual sections. Use the community vote above or visit All Sections to vote per section.

Read official text

Want the full experience?

Create a free account to cast verified votes on each section, get a personalized summary, and track your voting history.

Create Free Account