House lawmakers propose AI regulation bill that would override state laws for three years

New Bipartisan Legislation Takes a Big Step Forward in Restricting State Regulation of AI

A new draft bill that would regulate some of the most advanced artificial intelligence tools was released for public comment by Republican Rep. Jay Obernolte of California and Democratic Rep. Lori Trahan of Massachusetts on Thursday. The Obernolte-Trahan bill, dubbed the Great American AI Act, would ask large AI developers to keep the government informed about the development of its frontier models, make plans to mitigate the most severe harms to cybersecurity, and allow auditors to ensure those plans are followed. And while all of that may sound reasonable to people who want to make sure AI doesn't go all Skynet on us -- or just ensure it isn't used by rogue actors to dismantle important cyber or financial infrastructure -- the proposed federal legislation is controversial for one big reason: It would preempt some of the laws already passed at the state level. As Politico notes, the bill would also create the Center for AI Standards and Innovation (CAISI), an office in the Commerce Department's National Institute of Standards and Technology, to evaluate frontier models for the next three years. President Donald Trump's executive order on AI earlier this week sought to establish the office, but it has no funding without congressional approval. The law would allocate $300 million for the effort. "The threats AI poses to our national security, our safety, and our workforce are here and growing by the day," Rep. Trahan said in a statement posted online. "This bipartisan framework is designed to meet the challenges posed by this rapidly advancing technology without smothering American innovation," Trahan continued. "It protects workers, establishes real accountability for the most powerful frontier systems, and positions the United States to set the global standard on AI." Obernolte released a statement as well, emphasizing that it's important that any legislation on AI be bipartisan and stressing that input from other parties was welcome. "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," said Obernolte. "We are releasing this draft to hear from stakeholders, experts, and the public so we can strengthen the legislation before it is formally introduced." Other members of Congress from both parties are prepared to sign on to the draft bill, according to a press release, including Democrats Suhas Subramanyam of Virginia and Scott Peters of California, as well as Republicans Scott Franklin of Florida and Erin Houchin of Indiana. Elected Republicans have generally been more in favor of allowing AI companies to operate without any kind of government restraint, but Democrats and Republicans alike are still trying to understand what the best regulatory path forward might be in the face of threats to millions of jobs or massive security vulnerabilities. Sometimes, elected politicians play against their party's stereotype on AI. California Gov. Gavin Newsom, a Democrat, vetoed legislation that would've regulated AI in 2024 in a move widely seen as a way to protect OpenAI. And Florida Gov. Ron DeSantis, a Republican, has been a vocal critic of AI, even proposing a Citizen Bill of Rights for Artificial Intelligence late last year. On Thursday, the consumer advocacy organization Public Citizen called the proposed Great American AI Act a "disastrous proposal" that Big Tech is "celebrating." J.B. Branch, AI governance and technology policy counsel at Public Citizen, said the bill does nothing to address topics like algorithmic discrimination, consumer fraud, youth mental health harms, AI companions, and deepfakes. "At a time when states have led efforts to address AI-generated harms, Congress is proposing to take those tools away," said Branch. "Rather than establishing strong federal safeguards, the bill would preempt state action and defer to future federal testing, evaluation, and oversight frameworks that do not yet exist and will likely be impossible to pass in a divided Congress." The Tech Oversight Project also released new surveys Thursday of people who live in the districts of Rep. Obernolte (CA-23) and Rep. Trahan (MA-3), showing that majorities in both are skeptical of laws that might weaken state-level protections against AI. According to the group, 56% of Obernolte's constituents oppose the congressman's "efforts to weaken California's AI laws governing catastrophic risk." And 63% of Trahan's constituents feel the same way about local laws. "The Obernolte-Trahan bill replaces a state floor with a federal ceiling and trades away existing and future child safety, civil rights, and consumer protection laws for nothing in return," Sacha Haworth, executive director of The Tech Oversight Project, said in a statement to Gizmodo. Haworth argues that the proposed bill allows companies like OpenAI, Meta, Google, and Anthropic more power by requiring less oversight than local restrictions already passed in states like California and New York. And even those laws were passed with consultation from the AI companies, rendering them relatively weak. "Big Tech has a long, storied history of exploiting vague, opaque legislative language to evade transparency and accountability, and if the Obernolte-Trahan bill passes, it will grant Big Tech companies massive amnesty where there wasn't one before and rob the public of being able to use every tool possible to shape how AI affects our lives," said Haworth. "It is our sincere hope that Democrats and Republicans alike stop this AI amnesty bill from moving forward." Obernolte and Trahan are encouraging people to submit feedback on the proposed bill to the email address: [email protected].

What's inside the House draft bill to regulate AI

Why it matters: The draft's release is major step in what will be a difficult path forward for a bipartisan AI bill that can pass the House and the Senate. * The White House has been skeptical of any approach that imposes strict requirements on companies. Driving the news: Reps. Jay Obernolte (R-Calif.) and Lori Trahan (D-Mass.) rolled out a discussion draft of the Great American Artificial Intelligence Act on Thursday. * The draft comes days after President Trump signed an executive order on AI safety and cybersecurity. * Co-sponsors include Reps. Scott Franklin (R-Fla.), Suhas Subramanyam (D-Va.), Erin Houchin (R-Ind.), and Scott Peters (D-Calif.). What's inside: The 269-page framework would: * Preempt state laws on the development of AI models for three years. * Formally establish the Center for AI Standards and Innovation, tasked with making voluntary standards and guidelines, and appropriate $100 million per year from 2027-2029. * Require "large frontier developers" to write and implement plans addressing risks prior to releasing new models, as well as to report critical safety incidents to CAISI. The draft legislation would also protect AI whistleblowers and increase fines for AI-enabled fraud, as well as try to boost funding for AI literacy, education and research. * It would create a mechanism for the government to study AI's impact on the workforce and add some workforce protections around AI. * The draft touches on content moderation, cybersecurity, research security and international AI standards. What they're saying: "AI will shape our economy, workforce, national security, and daily lives for decades, and the framework governing it must be durable enough to survive changes in Congress, administrations, and political priorities," the lawmakers wrote in an op-ed. * "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." What we're watching: The lawmakers call the draft "the start of a serious national conversation" to get feedback from experts and the public ahead of the bill's formal introduction.

AI firms craft state rules as White House, Congress stall

Major artificial intelligence labs are done waiting on Washington to pass a national standard for AI, turning to state bills to carve out their own policy lines while Congress tries to catch up. Most AI labs support a national safety framework for AI that would eliminate the patchwork of state regulations, but they are also realistic about Washington's slow timeline and states' hunger to jump on the issue. While some firms are still set on fighting all state regulations, others like OpenAI and Anthropic are using them to their advantage to stake out their policy positions and inspire similar language at the federal level. OpenAI, the firm that birthed ChatGPT, has already seen success with multiple of its endorsed-state bills passing legislatures in three major blue states this year. Chris Lehane, OpenAI's chief of global affairs, told The Hill the company wants to create a "de-facto" national framework by cherry picking a handful of major AI bills. "A form of ... reverse federalism," Lehane, a former aide for former President Clinton, said in an interview with The Hill Tuesday. "You're basically getting the states to replicate each other." The Trump administration has spent more than a year pushing Congress to codify federal preemption of state AI laws but mixed feedback on the House's latest proposal, combined with separate negotiations in the Senate, indicate lawmakers still have a long way to go, with less than six months before a new Congress. "In a perfect world where we could wave a magic policy / political wand, you would have legislation passed at the federal level that would establish required safety standards," said Lehane. "As part of that, you have some type of very narrow preemption as it relates to those sort of catastrophic-type safety risks." Without a federal standard, Lehane said the company began discussing a state-focused approach about a year ago, since states are "clearly willing to put in some kind of safety standards." In the first half of this year, states across the country have introduced more than an estimated 1,500 bills addressing concerns around AI. These concerns have grown over the past year, with polling showing an increase in Americans' fears over artificial intelligence's impact on the job market, environment and national security. OpenAI focused on a handful of specific and somewhat similar proposals in states to determine whether the passage of one bill could influence the progress of others, Lehane explained. The company secured its latest win in Illinois, where state legislators last month passed SB 315, an AI safety bill some argue is the strongest yet on the issue. The bill, titled the Creates the Artificial Intelligence Safety Measures Act, requires large frontier developers to create, publish and annually update an AI framework with assessments on catastrophic risks, cybersecurity and more. SB 315 mirrors legislation passed in California and New York but goes a step further by mandating frontier labs with more than $500 million in revenue to submit third-party audits of their safety plans every year. OpenAI's endorsement came as a surprise to some in the tech industry, given past industry opposition to similar audit provisions, which were eventually eliminated in other bills. Explaining its reasoning, OpenAI said the bill "advances a risk-based approach focused on the most capable models and highest-consequent harms and because it further advances the emerging national framework for frontier AI safety." The bill heads to Illinois Gov. JB Pritzker's (D) desk, and the governor has indicated he will sign the measure. OpenAI also endorsed a bill in New York -- the Responsible AI Safety and Education (RAISE) Act, which was signed last year and similarly requires AI developers to publish their safety protocols and safety incidents. The company notably stopped short of fully endorsing and initially had concerns about California's AI safety bill, SB 53. The bill, signed by California Gov. Gavin Newsom (D) last fall, established new transparency requirements for large AI labs, including mandated safety frameworks and transparency and safety incident reporting. An OpenAI spokesperson said the firm had "productive conversations with California lawmakers about the bill from the company, adding the company was "happy with the outcome" and expressed its support for its signing. Lehane noted OpenAI looked for bills that included provisions explicitly acknowledging the new state artificial intelligence rules could be overruled by a federal standard. This language makes it a "little bit easier to ultimately, down the road, be able to put some of the political pieces in play to actually get that federal safety standard right," Lehane said, adding the three bills "have spoken to the idea of having a national safety standard." When asked whether OpenAI is considering red states for this strategy, Lehane said the organization is in conversations with Republican-led states, but "there's been less of an appetite." "Democratic states tend to be historically much more sort of active in the regulatory space," Lehane said. Meanwhile, Anthropic, one of OpenAI's top competitors, is the only other major AI lab heavily engaged on state bill endorsements, though this work began in 2024 -- far ahead of most AI companies' involvement -- and builds on a different strategy than OpenAI. The Claude creator has tried to set itself apart in the technology industry, emphasizing its safety-first approach to development and policies. Its ideology has put the company and its leadership at odds with the Trump administration's pro-innovation stance at times. Like OpenAI, Anthropic supports AI safety addressed at the federal level over a patchwork of state regulations but maintains "powerful AI advancements won't wait for consensus in Washington." In 2024, Anthropic was one of the only firms to speak positively in about California's Senate Bill 1047, which became a predecessor for the since-signed SB 53. CEO Dario Amodei said at the time the bill's "benefits likely outweigh its costs," while still calling the bill imperfect. Newsom eventually vetoed that sweeping bill, paving the way for California's new AI safety law: SB 1053. While its competitors pushed against it, Anthropic became the first AI company to endorse it, arguing it formalizes testing practices the firm and others already do. Like OpenAI, Anthropic also endorsed New York's RAISE Act, stating late last year the bill's safeguards are "raising the bar," along with Illinois's SB 315. An Anthropic spokesperson told The Hill the AI giant sees state endorsements as a "start" to the national conversation about AI regulation. In a blog post published Thursday, Anthropic acknowledged the flurry of state laws but noted "the rapid pace of acceleration means that transparency alone is no longer sufficient." Amodei and the company reiterated on Wednesday they believe Congress cannot preempt state law unless it passes a "strong" federal framework first. While these multibillion-dollar companies have the influence and sway to lobby state legislators and governors across the country, small and mid-sized technology firms are hoping they do not get left out of the picture. Graham Dufault, general counsel for the trade group Association for Competitive Technology, suggested this is playing out like privacy did at a state level in the absence of a federal framework. "As differing sets of requirements -- even if they look at first and at a high level - -- to be super similar, those differences across states, even if it's just disclosure requirements, will add up," Dufault told The Hill. Lehane maintained OpenAI has these small and mid-sized companies' interests in mind and has pushed for the scope of the state bills to only include the AI companies with a very high revenue threshold. Dufault expressed concerns that this is not enough. "Experience has shown that the scope always sort of inevitably inadvertently drags in smaller companies, either in indirect ways or direct ways," he said. Suzanne Borders, the CEO and founder of the immersive analytics platform BadVR, told The Hill she worries the major AI firms' state work could "give the appearance that they care," but does not address the end goal. "It ends up becoming a legal exposure headache for me," Borders said. "It keeps me up at night like am I answering the questions properly? Am I going to get sued?" "I don't know, but quite frankly, don't want to find out," she added.

US House lawmakers release draft bill to regulate AI

US ⁠Representatives ⁠Lori Trahan, a Democrat, and ⁠Jay Obernolte, a Republican, released draft ⁠legislation to regulate AI, according to a ⁠spokesperson for Trahan. A bipartisan pair of US House lawmakers released ⁠draft ⁠legislation on Thursday that would prohibit states from passing laws "targeting artificial intelligence model development," according to copy of the draft. The ⁠draft legislation, released by Democrat Lori Trahan ⁠of Massachusetts and Republican Jay Obernolte of California, would not bar states from regulating how AI technology is used, according to the draft. "We are releasing this draft to hear from stakeholders, experts, ⁠and the public so we can strengthen the legislation before it is formally introduced," Obernolte said in a statement.

House lawmakers introduce draft for national AI framework

A bipartisan pair of House lawmakers released a long-awaited draft of a national framework on artificial intelligence on Thursday, aiming to preempt some state laws on AI, minimize the technology's risks and expand research. The discussion draft, obtained by The Hill, proposes overriding state regulations that target AI model development for three years. This would not necessarily preempt state laws dictating how AI is used once released, according to the text. The draft "expressly does not preempt laws of general applicability, common law remedies, or laws regulating AI use or deployment," a summary of the draft stated. The prospect of an AI preemption law has been in talks for more than 18 months, but Washington has largely stalled on getting anything across the finish line. The Senate came close to passing a moratorium on new state AI regulations last year, but it fell through at the last minute. The proposal would also codify the Center of AI Standards and Innovation (CAISI), which was launched last June under the Commerce Department, would also be codified under the proposal. CAISI would oversee voluntary guidelines and standards for AI security, along with licensing independent verification organizations to audit labs' compliance. Larger frontier developers would also face new transparency requirements to publish a "frontier AI framework" including the technical and organization protocols it uses to evaluate and manage a model's "catastrophic risks." The labs would also be required to file a report with CAISI shortly after certain safety incidents, including those posing "imminent risk of death or serious injury." The two lawmakers also published an op-ed in Bloomberg Law on Thursday morning, emphasizing the discussion draft is not a "final product," but rather a way to start a broader conversation. Lawmakers have long wrestled with AI regulation that protects against the negative effects of the technology but also allows American companies to stay ahead of China in the so-called AI race. "Our framework includes safety and transparency requirements for frontier AI while avoiding a one-size-fits-all approach that would slow innovation," Reps. Jay Obernolte (R-Calif.) and Lori Trahan (D-Mass.) wrote Thursday. Scrutiny over AI's cybersecurity capabilities has drastically ramped up in Washington in recent months as backlash over the technology's impact on workforce and the environment grows among voters. The Trump administration has acknowledged these risks in recent months, but still largely promotes a hands-off approach to regulation. Trump and other White House tech leaders sent Congress an AI policy "wish list" earlier this year, with this framework serving as Congress's first real response to the White House's demands. It comes just days after President Trump signed an executive order laying out a voluntary government testing process for unreleased AI models. The order states AI labs can provide the government with their models for a testing period of up to 30 days before they plan to release them to the public. The order notably reduced the 90-day period for government testing after industry expressed concerns AI development works on a much faster timeline and 90 days could hamper competition with China and other labs. Obernolte and Trahan said Thursday said industry, safety experts and stakeholders say the order is an "important step," but "many are also asking whether our laws can keep up." "The question before Congress isn't whether AI will be governed," they wrote. "It's whether we will build a clear national framework that protects Americans, supports innovation, and ensures the US leads the world in shaping this technology." The lawmakers looked to quell concerns over AI's impact on the workforce, proposing the Secretary of Labor oversee data collection and hold workshops on the issue. The secretary would also be required to establish an AI Workforce Research Hub to conduct analysis of AI's workforce impact, scenario planning and policy recommendations. "We can't prepare Americans for the future if we're not measuring what's happening in real time," the lawmakers wrote. "Our proposal requires better federal data collection on AI's labor market impact, improved forecasting for occupations most likely to be affected, and additional transparency when AI is a substantial factor in qualifying mass layoffs."

House Members Propose Replacing State AI Laws With National Standard | PYMNTS.com

The discussion draft of the Great American AI Act is meant to encourage feedback from stakeholders, experts and the public before the bill is formally introduced, Reps. Jay Obernolte, R-Calif., and Lori Trahan, D-Mass., said in a Thursday (June 4) press release. Obernolte said in the release: "Artificial intelligence is advancing rapidly, which is why Congress must take a thoughtful and bipartisan approach to regulating this critical technology. 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." Trahan said in the release: "The threats AI poses to our national security, our safety and our workforce are here and growing by the day. This bipartisan framework is designed to meet the challenges posed by this rapidly advancing technology without smothering American innovation." According to a section-by-section summary of the bill released by the lawmakers, the bill would preempt any state or local law or regulation covering the development of AI models for three years. It would also require the comptroller general to submit a report on federal statutes and regulations that affect AI innovation or AI infrastructure. The bill would require transparency in frontier AI, independent verification organization audits and assessments, and anti-retaliation protection for AI whistleblowers. Other provisions address AI fraud deterrence, free speech, workforce education, labor market data, worker protections and adjustment assistance, cybersecurity, research and development, and international cooperation. In an op-ed published in Bloomberg Law, Obernolte and Trahan said that while several states have moved forward with AI legislation, risks created by AI are not bound by state lines. "Rather than allow protections to exist in only a handful of states or force innovators to navigate dozens of different legal regimes, our framework would establish one national standard," Obernolte and Trahan wrote. "That standard would extend core protections to every American while giving developers, researchers and businesses the clarity they need to invest and build responsibly." For all PYMNTS AI coverage, subscribe to the daily AI Newsletter.

US lawmakers propose 'Great American Act' to block state AI rules

* Read the draft text here. * Read the section-by-section summary here. A bipartisan group of US House lawmakers has released a draft bill that would bar states from regulating the development of AI models, reigniting a debate over whether the federal government or states should oversee. The discussion draft of the Great American AI Act, unveiled by Representatives Lori Trahan and Jay Obernolte along with other lawmakers, would bar states from passing laws that directly govern AI model development, including requirements for testing models before public release. States, however, would still be able to regulate how AI systems are used. Federal framework for AI: The proposal would create a national AI governance framework centred on a new Center for AI Standards and Innovation within the Commerce Department. The center would develop AI security standards, assess risks from advanced AI systems, coordinate with federal agencies and allies, and oversee audits of major AI developers. New obligations for AI companies: The draft also introduces transparency requirements for large frontier AI developers. Companies would have to publish safety frameworks, conduct risk assessments for advanced models, disclose key information before deployment, report critical safety incidents, and undergo independent verification. Violations could attract civil penalties of up to $1 million per violation. In a joint statement, the lawmakers said the bill aims to create a "national framework that protects Americans, supports innovation, and ensures the U.S. leads the world in shaping this technology." Pushback from consumer groups: The proposal has already drawn criticism from consumer advocacy groups like Public Citizen, which said the bill would leave "oversight largely to a federal government that has repeatedly failed to pass meaningful AI protections." The group also argued that the draft does not address issues such as algorithmic discrimination, consumer fraud, deepfakes, youth harms, and market concentration. The Information Technology Industry Council, which represents major technology companies, backed the effort, saying Congress should establish a national standard for AI development and deployment. Broader political context: The draft arrives as Congress continues to struggle to pass comprehensive AI legislation. It also follows recent moves by President Donald Trump's administration to pre-empt state AI rules. In March, the White House urged lawmakers to enact legislation overriding state AI regulations, and earlier this week Trump ordered leading AI developers to voluntarily submit their most advanced models for government cybersecurity testing before public release. Lawmakers have released the bill as a discussion draft and are seeking feedback from industry, researchers, and the public before formally introducing the legislation.