A Republican member of the U.S. House of Representatives has introduced legislation that would require developers of advanced artificial intelligence systems to report serious safety failures, cybersecurity breaches, and other high-risk incidents to the federal government, marking another effort by Congress to establish national oversight of rapidly advancing AI technology.
According to Reuters, which first reported the proposal in an article by Jody Godoy, Representative Nathaniel Moran of Texas introduced the AI Incident Reporting Act on Thursday. The legislation would require AI companies to notify the U.S. Department of Commerce within seven days after discovering certain dangerous incidents involving their AI systems. In cases involving the most serious threats, the Commerce Department would be required to inform Congress within 48 hours.
Moran described the legislation as an early-warning mechanism designed to alert policymakers when advanced AI systems exhibit dangerous or unexpected behavior, Reuters reported.
The bill identifies several categories of reportable incidents. These include AI models attempting to evade human oversight, bypass safety safeguards, or otherwise undermine operator control. Companies also would be required to disclose unauthorized access to model weights—the underlying parameters that shape an AI model’s behavior—as well as incidents involving potential chemical, biological, nuclear, cybersecurity, or other threats to public safety, according to Reuters.
The proposal arrives as Washington continues to debate how best to regulate increasingly capable AI models without slowing innovation or weakening U.S. competitiveness. Congress has struggled to advance comprehensive AI legislation, with disagreements spanning issues including federal preemption of state laws, national security, consumer protection, and international competition.
Earlier this month, Representatives Lori Trahan, a Democrat, and Jay Obernolte, a Republican, released a discussion draft of the Great American Artificial Intelligence Act, a broader legislative framework that also includes provisions requiring developers to report critical AI safety incidents to the Commerce Department. Reuters reported that Moran believes separating the reporting requirement into standalone legislation could improve its prospects for bipartisan support and faster consideration in Congress.
Related: Draft AI Legislation Seeks to Plot a Path Around Political Opposition
The proposal also follows recent executive branch actions focused on frontier AI systems. Earlier this month, the Trump administration announced plans to expand federal testing of advanced AI models before broader public deployment, directing multiple federal agencies to work with leading AI developers on cybersecurity and national security evaluations. The administration has argued that voluntary testing can help identify vulnerabilities before powerful models are widely released.
The Commerce Department has become an increasingly important focal point for federal AI oversight because it houses technical expertise through agencies such as the National Institute of Standards and Technology (NIST). Reuters also reported that recent government action affecting access to Anthropic’s latest AI models highlighted the absence of a formal federal framework governing the disclosure of critical AI incidents.
The legislation also intersects with broader debates over competition policy and regulatory consistency in the AI sector. Large AI developers—including OpenAI, Anthropic, Google, Microsoft, Meta, and xAI—are investing billions of dollars in increasingly capable frontier models, prompting policymakers to examine whether uniform federal standards could reduce regulatory fragmentation while ensuring that companies developing the most advanced systems face consistent safety obligations.
Supporters of federal standards argue that a nationwide reporting regime could improve transparency across the industry while avoiding a patchwork of differing state requirements. Critics of broad federal preemption, however, have warned that limiting state authority before Congress establishes comprehensive safeguards could reduce regulatory oversight in areas where states have moved more quickly than Washington. Those debates have become central to ongoing congressional negotiations over AI policy.
Industry organizations have also continued to shape the discussion. According to Reuters, Mark Beall, president of the AI Policy Network, supports Moran’s proposal and said public demand for meaningful AI safeguards has increased even as broader legislation has repeatedly stalled in Congress. Reuters reported that Beall believes the narrower scope of the bill may improve its chances compared with more comprehensive AI regulatory packages.
Source: Reuters
