The draft follows President Donald Trump's June 2 executive order directing federal agencies to create a voluntary framework for AI developers to provide the government early access for cybersecurity and national security review before release. The order is explicit: it does not create a mandatory licensing, permitting or preclearance process for AI development.

"Artificial intelligence is advancing rapidly, which is why Congress must take a thoughtful and bipartisan approach to regulating this critical technology," Obernolte said in announcing the discussion draft. 

"This discussion draft is an important step toward building a clear federal framework that promotes innovation, protects Americans from emerging risks, and ensures the United States continues to lead the world in AI," he added.

While the bill's regulatory provisions primarily target developers of advanced AI foundation models such as OpenAI and Anthropic, lawmakers also included a series of workforce and economic measures aimed at industries adopting AI technologies.

The proposal would require the U.S. Labor Dept. to establish an AI Workforce Research Hub, identify at least 15 occupations considered particularly sensitive to AI-driven disruption and publish recurring forecasts assessing the technology's effect on employment. Federal labor and business surveys would also be revised to collect more information on AI adoption and workforce impacts.

Employers conducting qualifying mass layoffs would be required to disclose when AI was a substantial factor in workforce reductions and estimate the percentage of job losses attributable to the technology.

The legislation arrives as the construction industry, a notoriously slow technology adopter, increasingly deploys AI tools for scheduling, estimating, document management, design support, safety monitoring and project analytics. AI development in construction is also progressing as other technologies adopted by the AEC industries before it such as CAD and BIM, by developing purpose-built language models with construction data separate from more general foundation models. 

While most within the construction sector—contractors, engineers and designers—would be affected as users rather than AI model developers, Associated General Contractors of America's 2025 Construction Hiring and Business Outlook survey found 44% of contractor respondents planned to increase investment in artificial intelligence in 2025.

AGC officials said contractors are already using AI to streamline back-office operations and support functions such as estimating, contract writing and marketing, while also applying the technology to analyze performance and safety data.

"A lot of the AI we are hearing about at present is related to streamlining back office operations, using AI tools to support estimating, contract writing and marketing services," Brian Turmail, vice president of association & industry image, says. "It is also a viable tool for analyzing performance and safety data."

The accelerating pace of AI evolution and fears of job obsolescence are the principal drivers of the legislation. "The threats AI poses to our national security, our safety, and our workforce are here and growing by the day," Trahan said.

“This bipartisan framework is designed to meet the challenges posed by this rapidly advancing technology without smothering American innovation," she added.

RELATED

AI in Construction: From More Work to Better Work

Data Centers, Infrastructure Also in Focus

The measure also touches on one of the fastest-growing sectors in construction: AI infrastructure.

AGC said one of its primary interests in the legislation is ensuring that AI-related investment continues to support demand for data-center construction projects.

"We appreciate that investments in AI technology are driving much of the significant new demand for data center projects, and want to make sure that demand remains strong," Turmail adds.

The bill would require the Government Accountability Office to identify federal statutes and regulations that "relate to infrastructure and energy and unduly burden artificial intelligence systems" and recommend legislative or administrative changes.

Another provision directs a federal review of liquid-cooling technologies used in AI data centers, which are drawing billions of dollars in investment as developers build facilities capable of supporting increasingly power-intensive computing workloads.

At the center of the proposal is the creation of the Center for AI Standards and Innovation within the Commerce Dept. 

The center would develop voluntary standards and best practices, evaluate advanced AI systems and support testing and assessment programs. The legislation would authorize $100 million annually from 2027 through 2029 for the center.

One of the proposal's most debated provisions would preempt state and local laws specifically regulating AI model development for three years following enactment. The draft preserves state authority over AI deployment and use, as well as generally applicable state laws and common-law remedies.

Critics argue the measure would go too far in limiting state oversight. Public Citizen said the proposal would "strip states of their authority" to address AI-related harms while relying on federal safeguards that have yet to be established. Both OpenAI, the maker of ChatGPT and Anthropic, the maker of Claude AI, have come out in support of the bill.

In a joint commentary accompanying the proposal, Obernolte and Trahan argued that AI systems are developed and deployed across state lines and require a consistent national framework.

"Rather than allow protections to exist only in a handful of states or force innovators to navigate dozens of different legal regimes, our framework would establish one national standard," they wrote.

The sponsors are seeking feedback from industry groups, labor organizations, researchers and other stakeholders ahead of formally introducing legislation.