What They’re Saying: Broad Coalition Lauds Bipartisan FRONTIER Act

Less than a week after Congresswoman Lori Trahan (D-MA-03) and Congressman Jay Obernolte (R-CA-23) introduced the Frontier Risk Oversight, National Transparency, Independent Evaluation, and Reporting (FRONTIER) Act , a broad coalition of stakeholders is praising the bipartisan legislation that would establish a federal risk-based framework governing the development and deployment of the most advanced artificial intelligence models. The FRONTIER Act would establish a tiered, risk-based national framework for oversight of the most advanced AI models, requiring the largest developers to publish safety frameworks, submit to independent third-party audits and verification, and report critical safety incidents to a new Under Secretary of Commerce for AI Security. It is an updated version of the Commerce-focused portions of the lawmakers’ Great American AI Act (GAAIA) discussion draft, reworked over seven weeks of stakeholder feedback to narrow its preemption of state law and sharpen its auditing and reporting provisions. “Not too long ago, a bipartisan bill this good and serious would have been a pipe dream. But things change quickly. This is by far the best AI safety bill ever introduced, in any jurisdiction,” said Charlie Bullock, Senior Research Fellow at the Institute for Law & AI . “There are still some flaws that I would like to see fixed, and some improvements to be made, but overall I'm very impressed with how well Reps. Trahan and Obernolte did on this. The next step is actually *passing* this kind of legislation, which unfortunately still seems like a pretty heavy lift. But this is a very promising first step.” “AI governance keeps running into the same wall. The technology is hard to measure and it moves faster than any law can keep up with,” said Andrew Freedman, Co-Founder and CEO of Fathom . “Trying to write the perfect rules and freezing them in place won’t work. What will work is a competitive market of independent verifiers who are accountable for real-world outcomes and who the government can actually count on. The FRONTIER Act shows we can move fast and still get this right.” “The latest version of the bipartisan Obernolte–Trahan framework, FRONTIER Act, is a serious step forward. It's one of the strongest efforts yet to establish a federal AI framework that protects Americans while preserving America's leadership in AI,” said the Alliance for a Better Future . “The bill gets several things right. Frontier developers would publish safety frameworks, file transparency reports before deployment, report serious safety incidents, and provide rapid notification when an imminent threat emerges. It also establishes meaningful emergency response authority and independent oversight for the most advanced AI systems. That's a targeted, risk-based approach that puts oversight where it's actually needed.” “Rep. Lori Trahan and Rep. Jay Obernolte's bipartisan FRONTIER Act deserves a lot of credit here; they are clearly engaged with stakeholders and civil society across the spectrum and have incorporated extensive feedback,” said Caleb Knapp, Senior Policy Manager and Secure AI Now . “In an ideal world, federal standards for frontier AI model pre-deployment would have crossed the finish line a year ago. Strongly addressing these issues is even more imperative as the technology develops further. We are seeing more instances of AI misalignment with OpenAI/Hugging Face, Kimi, and otherwise. Is the bill perfect? No. But it is the strongest federal frontier bill to date and builds on all the great work occurring at the state level in CA, NY, IL, and elsewhere.” “The FRONTIER Act is a meaningful bipartisan effort to strengthen oversight of the most advanced Al systems, and it's the strongest federal proposal to date. It builds on state regulations in many ways and incorporates a lot of the feedback that various stakeholders gave on the earlier discussion draft,” said the AI Verification & Evaluation Research Institute (AVERI) . “There are some remaining areas for improvement, including making sure that state laws are not preempted until their federal replacements are ready and clarifying some of the auditing provisions. But this is undoubtedly a step in the right direction and we look forward to working with Representatives Trahan and Obernolte, their staff, and other members of Congress to strengthen this proposal and advance effective oversight of frontier Al.” “Like a lot of people on [Twitter], I had some concerns regarding the initial federal framework discussion draft that Reps. Lori Trahan and Jay Obernolte released a couple weeks ago. I am happy to say that, after substantial engagement and feedback, they’ve crafted the strongest ai safety proposal to date in the FRONTIER Act. There’s still some improvements I’d like to see in the text, and I look forward to continued conversation with the sponsors. But! The narrower preemption and strong auditing provisions should inspire confidence that congress is evolving their approach quickly and worth engaging,” said Nat Purser, Director of U.S. Policy at AVERI . “The revised FRONTIER Act is meaningfully better than GAAIA, the earlier draft from Representative Trahan and Representative Obernolte. They deserve credit for engaging constructively with stakeholders and taking feedback seriously,” said Nathan Calvin, General Counsel Encode AI . “Overall, this is very much a step in the right direction. Some important issues remain, but we’ll continue engaging with the authors and appreciate the work they’ve put into improving the bill.” “It's a good bill—actually worth going for, I think,” said Anton Leicht, a Fellow with the Carnegie Endowment’s Technology and International Affairs team . “[Especially] after the past few weeks, federal third-party evaluation seems like a very good way to build capacity beyond exec branch scrambles fast, and to get actual insight into labs-as-entities and their internal deployments.” “The FRONTIER Act gets the AI regulatory balance right: focus transparency and oversight on the handful of companies pushing the frontier through a coherent federal standard, and leave the rest of the ecosystem alone. Rather than micromanaging model development, it builds audit-and-disclosure infrastructure so that Congress, courts, and the American people can stay informed. Independent verification organizations make the framework adaptive in a way static rules never could, letting oversight evolve as fast as the technology. This is what right-sized AI governance looks like: light-touch with dynamic, third-party oversight where the risks are greatest,” said Samuel Hammond, Chief Economist & AI Policy Director at the Foundation for American Innovation . “Seven weeks from discussion draft to introduced legislation is fast for Congress on anything; for AI, where the complaint is that they’ve been asleep at the wheel, it’s astounding. This wasn’t a panic bill scribbled over a weekend; the revision shows seven weeks of actual work, with gaps closed, clocks tightened, and new teeth added. The sponsor statements make it clear that they were watching the same news as the rest of us,” said Daniel Parshall, Ph.D., Founder and Director of the Canary Institute for AI Policy . A copy of the legislation introduced last week can be accessed HERE . A section-by-section summary can be accessed HERE . ###