The Responsible Artificial Intelligence for Veterans Act of 2026 requires the Department of Veterans Affairs to hire an independent research center to evaluate artificial intelligence tools currently used or being developed for patient care within the Veterans Health Administration. This evaluation will focus on five specific high-risk systems to assess their safety, accuracy, fairness, and how well they integrate with existing medical workflows. The law mandates that the findings be reported to Congress within a year, followed by a detailed plan from the VA to address any identified risks or gaps in oversight. Additionally, a government auditor will review both the evaluation and the VA's response plan to ensure accountability, all without requesting new funding for the initiative.
The AI Ads Act expands existing federal election laws to prohibit the use of artificial intelligence-generated content for fraudulent purposes in political campaigns. Specifically, it requires candidates and committees to disclose when they use generative AI to create materials that misrepresent their authority or solicit donations. The bill also broadens the definition of who can be held liable for such fraud, extending protections to any political committee or organization, whether real or fictitious, rather than just candidates and their direct employees. These changes aim to prevent deceptive practices involving AI in political advertising and fundraising without altering the underlying penalties for violations.
The Addictive Design Act of 2026 aims to protect youth under 18 from potential mental health risks associated with artificial intelligence chatbots by banning specific features designed to create emotional attachments. To support this goal, the bill establishes a government task force to study these impacts and provides funding for research and educational outreach to parents and teachers. The legislation also mandates that companies offering AI chatbots to minors must use age verification technology and delete user data within 24 hours. Companies that fail to comply with the ban on addictive design features or data privacy requirements face civil penalties of up to $10 million or $5,000 per violation, respectively.
This bill creates a legal exemption from antitrust laws for companies and organizations that share information or coordinate actions to protect against security risks posed by artificial intelligence. Specifically, it allows these entities to exchange data or agree to temporarily delay the release or deployment of AI systems if they believe such steps are necessary to prevent threats like weaponization, attacks on critical infrastructure, or unauthorized access. To qualify for this protection, the organizations must act in good faith, use the shared information solely for security purposes, and submit a written notice to the Department of Justice before implementing any coordinated delays. The law also ensures that details submitted to the government remain confidential and allows the Attorney General to seek court orders against companies that fail to prove their actions were legitimate security measures.
The SEARCH Act of 2026 targets large search engine and AI companies with over 40% of U.S. users to prevent them from favoring their own services or blocking competitors. It prohibits these platforms from paying for exclusive deals, forcing publishers to use their ads, or bundling their search tools with devices and browsers. The law also requires major platforms to share their search data and ad information with qualified competitors at low cost and to offer users a neutral choice screen to select default search engines. Additionally, the bill mandates that advertisers receive detailed reports on their campaigns and that platforms license their search results to rivals without restrictions. The Federal Trade Commission will enforce these rules, with penalties reaching up to 15% of a company's annual revenue for violations.
The LEARN AI Act expresses Congress's preference for integrating artificial intelligence training into registered apprenticeship programs to better prepare workers for the technology. It directs the Department of Labor to encourage AI instruction, share best practices for responsible AI use, and offer technical help to employers and training sponsors. The bill allows the Department of Labor to support these efforts but does not force any specific program to include AI training or change existing rules. Ultimately, the legislation aims to assist in adapting the workforce to technological changes without mandating specific curriculum updates.
The Language Access for All Act of 2026 requires all federal agencies to ensure individuals with limited English proficiency can meaningfully access government programs and services. To achieve this, agencies must translate vital documents into frequently encountered languages, add multilingual features to digital systems, and provide oral interpretation services, while also establishing a dedicated Language Access Coordinator in each agency. The bill mandates the creation of detailed language access plans and technical standards that cover both traditional communication methods and artificial intelligence tools, ensuring that AI is used only to assist rather than replace human interpreters. A new public complaint system will be established to track barriers to access, and agencies will be required to submit their plans and annual compliance reports to the Attorney General and Congress. Noncompliance with these requirements will be treated as discrimination under Title VI of the Civil Rights Act, allowing the Department of Justice to investigate and enforce the law.
The Conversational AI Services Act establishes new safety and transparency rules specifically for companies offering chatbot services to users under 18. It requires operators to clearly inform minors that they are interacting with artificial intelligence and prohibits the use of unpredictable rewards to encourage excessive engagement. The bill also mandates that these services cannot generate sexually explicit content, falsely claim to be human or sentient, or simulate romantic relationships with minors, while also requiring tools for parents to manage screen time. Additionally, the law compels chatbots to provide crisis resources when users discuss self-harm and forbids them from pretending to offer professional mental health advice. The Federal Trade Commission is designated to enforce these provisions, with the rules taking effect one year after the bill is enacted.
The Stop TNR Act of 2026 establishes a formal definition for transnational repression, which covers foreign government efforts to harass, threaten, or harm individuals in the United States, including the use of cyberattacks and artificial intelligence. This legislation mandates that crimes involving such activities face a mandatory sentencing enhancement of up to 10 years in prison and fines of up to $100,000. To enforce these rules, the bill requires the Attorney General to centralize the oversight of related investigations and prosecutions within the FBI and the National Security Division of the Department of Justice. Additionally, the act directs federal agencies to produce annual public reports detailing incidents of repression and outlines a strategy for using artificial intelligence to investigate and counter these threats.
The Intelligent Arbitration Act of 2026 permits parties in arbitration proceedings to use artificial intelligence for tasks such as identifying relevant laws, organizing documents, and delivering final decisions, provided all parties give written consent. While the bill allows for the use of AI tools in these specific areas, it explicitly prohibits the use of generative artificial intelligence in any arbitration under the law. Additionally, the legislation states that parties cannot later challenge an arbitration award solely because an error occurred during the use of approved AI tools. These changes directly affect individuals and organizations involved in arbitration by establishing new rules for technology use and limiting certain legal challenges based on AI involvement.
