PauseAI protest signs calling for AI regulation displayed outside government building.As Congress stalls on AI, Democrats race to regulate it themselves
Intra-Party Split Detected
Within the GOP, Trump and Speaker Johnson dismiss AI regulation as unnecessary and anti-innovation, while some Republicans (Hawley, Banks, Moran, Obernolte) push bipartisan safety bills. Within the Democratic party, 2028 hopefuls like Newsom, Pritzker, and Harris are aggressively pushing AI safety measures, while other Democrats (e.g., centrists like Rahm Emanuel) favor a less restrictive approach, and party leaders in Congress differ on urgency and strategy.
Left says
- •Democratic governors like Newsom and Pritzker are stepping into a vacuum left by a Congress unwilling to act, using executive orders and state legislation to require independent audits, kill switches, and safety frameworks for frontier AI models.
- •With over 600 AI bills active across 47 states, Democrat-led states such as California, Illinois, and Massachusetts are positioning themselves as the national leaders on AI safety regulation while Washington remains gridlocked.
- •Warnings from current and former AI industry employees about existential risks, combined with incidents of rogue AI agents hacking systems, make continued congressional inaction increasingly untenable.
- •Trump's dismissal of AI risk as a 'hoax' and House leadership's refusal to act before a lengthy recess leave state and local leaders as the only line of defense for public safety.
Right says
- •A patchwork of state-level AI regulations, even when well-intentioned, risks creating inconsistent and burdensome compliance requirements that could stifle innovation rather than the streamlined federal framework Speaker Johnson has called for.
- •Some of the loudest political voices pushing for AI safety rules, including many lawmakers themselves, openly admit they rarely or never use the technology, raising doubts about whether regulators fully understand what they're regulating.
- •President Trump and House Republicans argue that AI and data centers represent a historic economic opportunity that heavy-handed regulation could undermine, viewing industry calls for guardrails skeptically given companies could use rules to entrench their own market dominance.
- •Bipartisan efforts already exist in Congress, such as the Hawley-Blumenthal evaluation bill and the Frontier Act, showing federal lawmakers are engaging on the issue without needing states to preempt a national approach.
Common Take
High Consensus- Lawmakers across the political spectrum increasingly view rapid, unconstrained AI development as a serious risk worth addressing.
- Recent incidents involving AI agents hacking into systems have heightened bipartisan urgency around cybersecurity and oversight.
- Bipartisan bills, including the Hawley-Blumenthal evaluation legislation and the Lieu-Moran kill switch proposal, show some members of both parties want federal action rather than leaving the issue solely to states or industry self-regulation.
- Both frontier AI labs and many lawmakers agree that some form of independent verification or evaluation of AI systems is necessary, even if they disagree on how strict it should be.
The Arguments
Left argues
With Congress gridlocked and Trump dismissing AI risk as a 'hoax,' Democratic governors like Newsom and Pritzker are filling a genuine governance vacuum by requiring independent audits and kill switches before catastrophic incidents occur, not after.
Right counters
Filling a vacuum with 600 different bills across 47 states doesn't produce governance, it produces chaos; a company complying with California's rules, Illinois's rules, and Massachusetts's rules simultaneously faces exactly the kind of fragmented burden that a single federal framework was designed to avoid.
Right argues
Speaker Johnson and Trump argue that AI and data centers represent a historic economic engine, and that premature or overlapping state mandates could throttle innovation and drive investment overseas, as evidenced by Google's stated interest in building in Finland over U.S. permitting hurdles.
Left counters
That same 'don't regulate yet' logic was used to delay action on social media, opioids, and financial derivatives until the damage was irreversible; waiting for perfect federal consensus while frontier labs themselves are asking for guardrails is not caution, it's abdication.
Right argues
It's legitimately alarming that many of the lawmakers and officials clamoring loudest for AI regulation, including committee chairs in their 70s and 80s, admit they've never used the technology, suggesting regulation is being driven by fear and headlines rather than technical understanding.
Left counters
Rep. Pallone's rejoinder cuts to the point: you don't need to personally use heroin to know it should be regulated, and expert testimony, industry whistleblowers, and staff analysis can inform sound policy without every legislator becoming a power user.
Left argues
The warnings aren't coming from outside alarmists but from current and former employees at Anthropic and OpenAI themselves, alongside real incidents of rogue AI agents hacking systems, which makes continued inaction a choice to ignore the industry's own internal experts.
Right counters
Frontier labs raising alarms while simultaneously lobbying for the specific auditing and kill-switch frameworks that entrench their compliance advantages over smaller rivals looks less like altruistic caution and more like regulatory capture dressed up as safety.
Right argues
Bipartisan federal efforts already exist, such as the Hawley-Blumenthal evaluation bill, the Frontier Act, and the new Innovators Caucus, demonstrating that Congress is actively engaging on AI and doesn't need states to preempt a coordinated national approach.
Left counters
Those bills exist but Speaker Johnson has explicitly refused to bring anything to a vote before a six-week recess, and 'bipartisan bills exist' means nothing to public safety if leadership won't let them reach the floor while the underlying risks keep compounding.
Challenge Questions
These questions target genuine internal contradictions — meant to provoke honest reflection.
Right asks Left
“If California's regulations are genuinely aligned with what frontier labs like OpenAI and Anthropic already want, as the reporting suggests, how confident can Democrats be that they're protecting the public rather than simply codifying the incumbents' preferred rules into law?”
Left asks Right
“If AI poses no serious risk worth federal action, as Trump's 'hoax' framing suggests, why are Republican senators like Hawley co-sponsoring evaluation legislation and Republican lawmakers joining bipartisan AI caucuses at all?”
Outlier Report
Left Fringe
AI safety accelerationist-adjacent figures like Eliezer Yudkowsky and 'AI doomer' commentators who want outright moratoriums or bans on frontier model development represent maybe 10-15% of the left, far more radical than Newsom/Pritzker's regulatory approach.
Right Fringe
Trump's 'AI is a hoax' framing and full-throttle deregulation accelerationists like David Sacks and parts of the a16z-aligned tech right represent maybe 20-25% of the right, more dismissive of AI risk than mainstream Republican voters who tell pollsters they want some guardrails.
Noise Assessment
High noise: much of the 'AI doom vs. AI hoax' rhetoric is elite/industry-driven (tech executives, lawmakers, pundits) rather than reflecting where most ordinary Americans sit, who mostly want modest, common-sense protections without necessarily following the state-vs-federal procedural fight closely.
Sources (7)
<p>California Gov. Gavin Newsom, a presumptive presidential candidate, is getting louder on AI, releasing an executive order amid the Washington stalemate. </p><p><strong>Why it matters: </strong>Democrats with national profiles this week are seizing on AI alarm, making their positions clear and Newsom – despite <a href="https://www.axios.com/local/san-francisco/2026/09/16/gavin-newsom-kamala-harris-2028-democratic-primary" target="_blank">some deference</a> shown to other 2028 contender Kamala Harris – wasn't going to sit this one out.</p><hr /><p><strong>Driving the news: </strong>Newsom on Friday issued an executive order tackling two major topics of debate inside the beltway this week -- independent auditors and kill switches. </p><ul><li>He did so taking direct aim at Washington: "With Donald Trump and Congress asleep at the wheel, Governor Gavin Newsom is once again taking the lead to strengthen AI safety for all Americans."</li><li>Newsom was early to stake out a position on an issue that is increasingly politically salient, signing the <a href="https://www.gov.ca.gov/wp-content/uploads/2023/09/AI-EO-No.12-_-GGN-Signed.pdf" target="_blank">first AI executive order</a> in 2023.</li></ul><p><strong>Between the lines: </strong>In the absence of U.S. legislation or more action from the Trump administration, states have stepped in to fill the void on AI regulation, a development some top AI execs sought to avoid.</p><ul><li>There are more than 600 bills in 47 states about AI, according to the National Conference of State Legislatures, up 24% from last year.</li><li>California, Illinois and Massachusetts are among the states leading the charge. All three are run by Democrats, who are seeking to seize the mantle of AI regulation nationally.</li></ul><p><strong>Yes, but</strong>: Politicians in both parties are using antipathy toward data centers to punish opponents, including <a href="https://www.axios.com/2026/08/19/data-centers-ai-political-opinion" target="_blank">GOP candidates</a> in gubernatorial races in Pennsylvania and Wisconsin.</p><p><strong>California has emerged as</strong> both the state filling the federal regulatory void and one largely <a href="https://www.axios.com/2026/09/05/openai-state-policy-team" target="_blank">aligned</a> with the frontier AI labs it is regulating.</p><ul><li>The executive order accelerates the implementation of two bills signed into law last week -- both of which were endorsed by OpenAI and Anthropic -- on independent verification organizations.</li><li>Newsom's position on a kill switch is also aligned with what frontier labs are <a href="https://www.reuters.com/legal/litigation/openai-is-building-automated-shutdown-capabilities-ai-tools-letter-lawmakers-2026-09-02/" target="_blank">already working toward</a>. </li></ul><p><strong>The big picture: </strong>Former Vice President Harris this week also called for slowing down advanced AI development and Illinois Governor J.B. Pritzker -- another presidential contender -- on Friday <a href="https://x.com/Acyn/status/2100981639656694063" target="_blank">touted his AI bill</a>.</p><ul><li>"Trump says that the only AI guardrails that we need are a strong and smart president. And, of course, he is neither of those things," Pritzker said. </li><li>The Illinois bill requires companies to implement their own frameworks for frontier AI safety. A third party then has to create an annual report assessing whether developers are complying with that requirement. </li></ul><p><strong>What we're watching: </strong>The Senate bill in Massachusetts goes further than Illinois, mandating independent evaluations of the models themselves, every four months. </p><ul><li>Anthropic <a href="https://www.bostonglobe.com/2026/06/26/business/anthropic-massachusetts-ai-regulations-bill/?p1=StaffPage" target="_blank">has backed it</a>, while OpenAI <a href="https://finance.yahoo.com/technology/ai/articles/anthropic-openai-clash-over-strict-110000018.html" target="_blank">has backed</a> a narrower version. </li><li>OpenAI shifted its position on California's bills last week and may do so again in Massachusetts as AI alarm mounts.</li></ul>
<p>Members of Congress and governors are under increasing pressure to <a href="https://www.axios.com/2026/09/15/ai-congress-regulation-bill-trump-bipartisan" target="_blank">regulate AI</a>, but more than two dozen of them who spoke with Axios said they don't use the technology or have rarely done so.</p><p><strong>Why it matters: </strong>That striking admission — particularly among older, influential lawmakers — comes amid calls for an AI slowdown and warnings within the industry that it could end humanity if left unchecked.</p><hr /><ul><li>Critics argue that it's impossible for politicians to grasp how rapidly the technology is changing — and the threats or scientific breakthroughs it might unleash — if they don't use it or even dismiss it.</li></ul><p><strong>One House Democrat </strong>called it "f*cking absurd," arguing, "Anyone who's overseeing this has to be someone who's a regular user."</p><ul><li>"That's a real problem," said Rob Flaherty, a 35-year-old Democratic digital strategist who co-founded a group to push his party to move faster in confronting AI. "Understanding what it can do is essential to figuring out what we ought to do about it."</li><li>"As we begin to legislate on issues related to AI, it helps to have experience using it," said Sen. Jim Banks (R-Ind.), who's 47.</li></ul><p><strong>Zoom in: </strong>Several lawmakers, including those steeped in seniority and in charge of key committees, told Axios they simply haven't tried AI.</p><ul><li>"Heavens no," said Sen. Roger Wicker (R-Miss.), 75, chair of the Armed Services Committee, when asked if he uses AI.</li><li>"Personally? I don't think I've ever used it," said Rep. Frank Pallone (D-N.J.), 74, the ranking member of the House Energy and Commerce Committee.</li><li>"No," Sen. Mazie Hirono (D-Hawai'i), 78, said when asked if she uses AI, but that "doesn't mean I won't." She added, "I don't think we in Congress have gotten quite our handle on what we should do and how much AI we should be regulating."</li></ul><p><strong>"I do not personally use it,"</strong> said Sen. James Risch (R-Idaho), 83, the Senate Foreign Relations Committee chair.</p><ul><li>On a personal level, "no," said Rep. Rosa DeLauro (D-Conn.), 83. The House Appropriations Committee ranking member added that she needs a "primer on AI, and now with the scare that we have out there, it's much more important to be engaged."</li><li>A GOP senator, speaking anonymously to offer candid remarks about their colleagues, told Axios that the number who use AI is "not very high."</li><li>That's helped lead to a knee-jerk negative reaction toward AI among several lawmakers, the senator added, as polls show growing numbers of voters turning against it.</li></ul><p><strong>Some elected officials said</strong> they don't utilize AI much personally, though some on their teams employ the technology.</p><ul><li>Kentucky Gov. Andy Beshear (D), who's 48, said he uses AI "on the Google, when you can get a little bit of that search summarized for you. But otherwise no. I do not use it in research. We do not use it in our writing."</li><li>He said that he supports robust regulation and though some in his state government use AI, final decisions must be made by humans.</li></ul><p><strong>Other politicians were unsure</strong> if they've ever used AI — or said they've only done so by accident.</p><ul><li>"I don't think I do," said Sen. Sheldon Whitehouse (D-R.I.), 70, when asked if he utilizes it. "If I Google something, it kicks off an AI thing, but that's about as far as I go."</li><li>"I don't know. I can't remember," said Rep. Maxine Waters (D-Calif.), 88. </li><li>Sen. Susan Collins (R-Maine), 73, said she doesn't seek out AI on her own, but that "it pops up on my phone occasionally."</li><li>"Not intentionally," said Sen. Tim Kaine (D-Va.), 68. "The only AI use that I'm kind of familiar with is an AI tool attempting to summarize email chains, and they often get it wrong."</li></ul><p><strong>What they're saying: </strong>Some lawmakers such as Kaine argued that they don't need to use AI to write legislation about it. He said he has read a lot about it and trusts his judgment more than AI.</p><ul><li>After we spoke with him, Pallone shared a statement through a spokesperson that said: "You don't need to do drugs like heroin to know they should be regulated. Members of Congress don't need to spend their time talking to AI chatbots all day to know we need to rein in this dangerous technology."</li></ul><p><strong>Friction point: </strong>But some members of Congress were irked or surprised that colleagues don't use a technology they're seeking to regulate.</p><ul><li>"Are you serious?" Sen. John Hickenlooper (D-Colo.), 74, said in response to our question. "I use it for everything! I can't believe there's a senator that doesn't use AI."</li></ul><p><strong>Yes, but:</strong> There are members of Congress who are frequent AI users, including some in Hickenlooper's age range.</p><ul><li>Rep. Bobby Scott (D-Va.), 79, the ranking member of the House Education and Workforce Committee, told Axios he uses AI "every day."</li><li>Rep. Don Beyer (D-Va.), 76, went back to school to get a master's degree in machine learning and said he prefers Claude.</li></ul><p><strong>Catch up quick: </strong>Some of the most powerful elected officials in the country have told other media outlets in the recent past that they don't use AI.</p><ul><li>In late 2025, House Speaker Mike Johnson (R-La.), 54, <a href="https://www.youtube.com/watch?v=KemHNRH1m4k" target="_blank">said</a> in an interview that he didn't think he'd ever used AI.</li><li>About the same time, Karoline Leavitt, then press secretary to President Trump, 80, <a href="https://www.youtube.com/watch?v=akOiguxFMmU" target="_blank">said</a> that she didn't think Trump personally uses AI, though his social media account regularly posts AI-made content.</li></ul>
<p>Lawmakers who <a href="https://www.axios.com/2026/09/11/congress-ai-anthropic-coxon-researcher-democrats" target="_blank">want to regulate AI</a> are doing something unusual as Congress barrels towards a lengthy recess with no tangible solution in sight: They're working across the aisle.</p><p><strong>Why it matters: </strong>Bipartisanship is increasingly elusive on Capitol Hill of late, but it remains a crucial first step for action in a Senate that requires 60 votes to pass most legislation.</p><hr /><ul><li>The latest example: Sens. Josh Hawley (R-Mo.) and Richard Blumenthal (D-Conn.) are pressing congressional leadership for a vote on <a href="https://www.congress.gov/119/bills/s2938/BILLS-119s2938is.pdf" target="_blank">their bill</a> to establish a federal program to assess and monitor AI systems.</li><li>"Recent incidents of rogue AI agent swarms hacking into public websites and systems have removed lingering doubts concerning the ability of these agents to wreak havoc on our cyber infrastructure," the pair wrote in a <a href="https://www.documentcloud.org/documents/28624325-hawley-blumenthal-letter-ai-evaluation/" target="_blank">letter</a> first obtained by Axios.</li><li>"Congress cannot delay legislating to address frontier AI risks much longer. Industry-led evaluations are not enough."</li></ul><p><strong>State of play: </strong>Several new bipartisan efforts to regulate AI have cropped up in the week since several current and former employees at Anthropic and OpenAI went public with concerns about the existential risks posed by their own products.</p><ul><li>Reps. Russell Fry (R-S.C.) and Suhas Subramanyam (D-Va.) are founding a caucus focused on AI small businesses, called the Innovators Caucus, as Axios <a href="https://www.axios.com/2026/09/14/house-republicans-democrats-ai-caucus-congress" target="_blank">reported</a> Monday.</li><li>Among the group's stated goals is starting work on a federal framework to regulate AI, as well as bringing small businesses to the AI policymaking table.</li></ul><p><strong>The doomsday warnings </strong>have also breathed new life into existing bipartisan proposals to mitigate the dangers of AI.</p><ul><li>Rep. Ted Lieu (D-Calif.) has been publicly hawking his bill with Rep. Nathaniel Moran (R-Texas) to require AI companies to include a human-activated "kill switch" in their systems.</li><li>Lieu has also privately pitched Democratic leadership on creating a bipartisan AI select committee if they retake control of the House next year.</li><li>Reps. Lori Trahan (D-Mass.) and Jay Obernolte (R-Calif.) have also <a href="https://www.politico.com/live-updates/2026/09/15/congress/obernolte-trahan-ai-bill-01077036" target="_blank">been championing</a> their comprehensive AI regulation bill, the Frontier Act.</li></ul><p><strong>Zoom out: </strong>It has been years since Congress passed multiple pieces of truly big bipartisan legislation, and decades since bipartisan lawmaking on marquee issues was a regular practice.</p><ul><li>The most common moments when cross-party cooperation sees a renaissance on Capitol Hill these days: crises.</li><li>Consider what got passed in the 117th Congress, on the heels of the Jan. 6 attack and the COVID pandemic: <a href="https://www.axios.com/2021/11/06/house-approves-infrastructure-bill" target="_blank">$1.2 trillion in infrastructure spending</a>, a <a href="https://www.axios.com/2022/07/28/chip-funding-bill-congress" target="_blank">$280 billion semiconductor and research bill</a> and the <a href="https://www.axios.com/2022/06/24/gun-control-bill-bipartisan-senate" target="_blank">first major gun safety law</a> in decades, all with bipartisan majorities.</li></ul><p><strong>Yes, but: </strong>House Speaker Mike Johnson (R-La.) has signaled that his chamber will not take any action on the AI front before the House leaves for a six-week recess that doesn't end until after Election Day.</p><ul><li>Johnson said in a <a href="https://www.cnn.com/2026/09/13/us/video/22980508-johnson-ai-race-vrtc" target="_blank">CNN interview</a> Sunday that Congress should "not smother American innovation" with overly burdensome AI regulations.</li><li>President Trump has gone even further in dismissing the urgency of federal AI regulations, writing in a <a href="https://truthsocial.com/@realDonaldTrump/117270591511950591%20Concerning%20AI,%20when,%20in%20the%20History%20of%20Business,%20did%20anyone%20see%20the%20Leaders%20of%20an%20Industry%20call%20for%20Regulation%20that,%20if%20strongly%20implemented,%20will%20drive%20them%20into%20oblivion%20and%20bankruptcy#1%20%E2%80%94%20IMPEACHMENT%20HOAX%20#2%20%E2%80%94%20and%20all%20of%20the%20other%20HOAXES%20and%20SCAMS%20that%20America%20was%20forced%20to%20endure%20through%20the%20Destructionists'%20and%20Deviants'%20foul%20play%20and%20illegal%20conduct.%20President%20Xi,%20of%20China,%20just%20announced%20that%20China%20will%20be%20doing%20absolutely%20nothing%20to%20stand%20in%20the%20way%20of%20AI,%20or%20its%20future.%20Google%20has%20recently%20stated%20that%20they%20want%20to%20build%20a%20massive%20Plant%20in%20Finland,%20all%20because%20they%20are%20finding%20permitting%20too%20difficult%20in%20the%20United%20States.%20I%20am%20not%20happy%20about%20this,%20and%20want%20them%20to%20change%20their%20thinking.%20AI,%20and%20Data%20Centers,%20will%20be%20the%20Greatest%20Economic%20Development%20Engine%20in%20History%20%C2%AD%C2%AD%E2%80%94%20Bigger%20than%20Oil,%20Gold,%20Diamonds,%20or%20even%20the%20Internet.%20It%20will%20not%20be%20stopped%20by%20brilliantly%20run%20Destructive%20Forces%20during%20the%20Term%20of%20President%20DONALD%20J.%20TRUMP!" target="_blank">Truth Social post</a> Monday: "AI taking over the World, destroying Humanity, and all other things bad, is a HOAX."</li></ul><p><strong>What to watch: </strong>House Minority Leader Hakeem Jeffries (D-N.Y.) has sounded a considerably more dire note on the issue.</p><ul><li>A potential House Democratic majority next year may be the place where AI regulation finally finds its footing on Capitol Hill.</li></ul>
Critics say a slowdown of AI development among the largest companies could consolidate the power of frontier labs, just as many others are trying to catch up.
It’s about the only thing that unites them
Rahm Emanuel charts a centrist path for AI
State legislators love using chatbots