Representatives Jay Obernolte of California and Lori Trahan of Massachusetts unveiled the Great American Artificial Intelligence Act on Wednesday, a sweeping 269-page discussion draft that represents the most ambitious attempt yet by Congress to establish a comprehensive federal framework for regulating artificial intelligence. The bipartisan bill seeks to balance innovation with safety by creating new obligations for large frontier AI developers while temporarily blocking states from enacting their own AI development regulations. The legislation was released as a discussion draft, signaling that its sponsors are seeking input from stakeholders before moving toward formal introduction.
The centerpiece of the bill is a three-year preemption of state laws related to AI development, a provision designed to prevent a patchwork of conflicting regulations that technology companies argue would stifle innovation and create compliance nightmares. Under the proposed framework, states would retain authority to regulate the use and deployment of AI systems after a model has been released, including in areas such as civil rights, labor protections, copyright enforcement, child safety, and data privacy. The preemption provisions would automatically expire after three years, at which point Congress would need to reassess the regulatory landscape.
Large frontier AI developers would face significant new obligations under the legislation. Companies meeting certain computational thresholds would be required to publish comprehensive risk management plans detailing how they identify, assess, and mitigate potential harms from their AI systems. These developers would also be obligated to report safety incidents to federal regulators and submit to regular third-party audits of their safety practices. The bill establishes civil penalties and the possibility of injunctive relief for companies that fail to comply with these requirements.
The legislation arrives amid a surge of AI regulatory activity at the state level, with more than 300 AI-related bills currently progressing through state legislatures across the country. Proponents of the federal bill argue that this fragmented approach creates uncertainty for developers and could ultimately harm American competitiveness in the global AI race. They contend that a unified national standard would provide clearer guidance while still protecting consumers and addressing legitimate safety concerns about increasingly powerful AI systems.
Civil liberties organizations and consumer advocacy groups have voiced strong opposition to the preemption provisions. The American Civil Liberties Union issued a statement criticizing the bill for potentially preventing states from protecting their residents against AI-related harms during the three-year preemption period. Public Knowledge, a digital rights organization, similarly argued that the bill prioritizes industry preferences over public safety. Both organizations expressed concern that the temporary nature of the preemption could become permanent through successive extensions.
The bipartisan sponsorship of the bill reflects an unusual moment of cross-party agreement on the need for federal AI governance, even as significant disagreements remain about the specifics. Obernolte, a Republican who chairs a congressional AI task force, has emphasized the importance of maintaining American leadership in AI development. Trahan, a Democrat, has focused on ensuring that regulatory frameworks include meaningful accountability mechanisms and protections for vulnerable populations. Their collaboration suggests that AI regulation may represent a rare area where legislative compromise is possible in an otherwise polarized Congress.
Industry reaction to the draft has been cautiously positive, with major technology companies expressing support for the concept of federal preemption while raising questions about specific compliance requirements and computational thresholds. Looking ahead, the bill faces a lengthy legislative process that will include committee hearings, stakeholder consultations, and likely significant revisions before any floor vote. Congressional observers note that even if the bill does not pass in its current form, it establishes an important baseline for future AI legislation and signals that federal regulation of the technology sector is increasingly viewed as inevitable by lawmakers on both sides of the aisle.
Comments