Democrats and Republicans Want to Regulate Artificial Intelligence, But They Cannot Agree on How to Do It
AI regulation has become one of the most heated debates in the U.S. Congress — and for good reason. For years, the tech sector grew with virtually no significant guardrails, and now lawmakers are trying to figure out how to bring some order to the house before the problems outgrow any possible solution. Public pressure has ramped up considerably, cases of misuse have multiplied, and members of Congress simply cannot look the other way anymore.
Democrats and Republicans agree that something needs to be done, but that is exactly where the conversation starts to fall apart. Each party carries a very different vision about what to regulate, how to regulate, and most importantly, who should be held accountable when things go wrong. These differences are not minor — they reflect completely distinct philosophies about the role of government in the economy and in people’s everyday lives.
At the federal level, bills written by Republicans tend to be less concerned with policing how individuals use the technology and more focused on regulating the development and deployment of the underlying technology — the large language models, or LLMs. Bills written by Democrats, on the other hand, typically focus on the harms caused by individual use of these tools, such as deepfakes, image exploitation, fraud, and large-scale disinformation. For them, the problem is not necessarily the model itself, but rather how certain people choose to use it in harmful ways.
Sounds simple, right? But when you dig into the bills that are actually making their way through Congress right now, things get a lot more complicated. Some of these bills have real potential to protect people by creating legal mechanisms that hold companies and individuals accountable for concrete harms. Others, however, could slow down one of the biggest technological revolutions in history — and might not even protect anyone in the end. 😬
The Democratic Side: Focus on Deepfakes and Misuse
A classic example of the Democratic approach came from Senator Amy Klobuchar of Minnesota. She was so outraged by a deepfake someone created of her — a video that, it should be noted, went viral across the internet — that she made a public appeal to Congress to affirm the right of individuals to demand that social media platforms remove deepfakes that use their voices and likenesses. In California, Democratic Governor Gavin Newsom signed three bills in 2024 that restricted the use of AI to create political content considered deceptive ahead of elections.
On the federal legislative front, Senator Dick Durbin, a Democrat from Illinois, introduced the DEFIANCE Act, which passed the Senate by unanimous consent in January and is being championed in the House by Representative Alexandria Ocasio-Cortez. This bill would make it a civil offense to create digital forgeries depicting intimate activity or nudity. It is a direct response to the explosive growth of non-consensual sexually explicit deepfakes, a problem that disproportionately affects women and minors.
While the DEFIANCE Act does not impose liability on AI companies for misuse by individuals, another Durbin bill goes in that direction. The AI LEAD Act, introduced in September and co-sponsored exclusively by Republican Senator Josh Hawley, would make AI developers and deployers liable when a user’s application of a system causes harm. Hawley frames the AI LEAD Act as a tool to empower parents to sue tech companies when AI products harm their children. However, as critics point out, virtually any product imaginable can be used maliciously. It would be like holding car manufacturers liable for every accident caused by reckless drivers.
The Republican Side: Control Over Models and Their Data
On the other side of the political aisle, Senator Josh Hawley, a Republican from Missouri, has been one of the most active — and most controversial — voices in the AI regulation debate. Hawley does not just want to ban self-driving cars to protect unionized truck drivers or bar minors from accessing AI chatbots. He wants frontier AI developers to submit their models to the Department of Energy for potential nationalization before receiving permission to deploy them commercially.
Hawley’s AI Accountability and Personal Data Protection Act prohibits the use of copyrighted materials that were legally acquired for AI training without permission from the rights holder. This bill is co-sponsored by Democratic Senators Richard Blumenthal of Connecticut and Peter Welch of Vermont. The proposal appears to be a response to the Bartz v. Anthropic case, in which the court determined that Anthropic did not violate the Copyright Act by training its LLM on legally acquired copyrighted works. However, Anthropic was found liable for copyright infringement for using more than 7 million copies of copyrighted books obtained illegally from pirate sites. If passed, this bill could seriously harm AI developers, who rely on legally acquired public and private data to train increasingly sophisticated models.
Receive the best innovation content in your email.
All the news, tips, trends, and resources you're looking for, delivered to your inbox.
By subscribing to the newsletter, you agree to receive communications from Método Viral. We are committed to always protecting and respecting your privacy.
Another Hawley bill, the Artificial Intelligence Risk Evaluation Act, also co-sponsored by Blumenthal, would require developers to hand over detailed information about their frontier LLMs to the Department of Energy before deployment. The department would then assess whether various adversarial scenarios are likely. If it concluded they were, it would have authorization to nationalize the technology. Think about the impact of that: fewer people will want to push the technological frontier if the government has the right to seize any company whose product is deemed too good.
The GUARD Act and Protecting Minors
Among all the bills under discussion, Hawley’s GUARD Act has the best shot at becoming law. It boasts no fewer than 12 co-sponsors from both parties, including names like Katie Britt, Tom Cotton, Ruben Gallego, Maggie Hassan, Mark Kelly, James Lankford, Mike Lee, Chris Murphy, Mark Warner, and Catherine Cortez Masto. The legislation would not only prohibit chatbots from producing sexually explicit content for minors — it would completely ban the provision of any AI companion to minors.
To comply with this sweeping regulation, chatbot companies would be required to freeze all user accounts, only unlocking them after users provide verifiable age data through a reasonable age-verification process. These processes include providing a government-issued ID or biometric data to AI companies. As the Electronic Frontier Foundation has warned, this means every interaction with a chatbot could be tied to a user’s verified identity — a serious privacy concern.
And this risk is not theoretical. AU10TIX, an identity verification software used by TikTok, Uber, and X, recently left personally identifiable information exposed for over a year. Imagine the potential damage if biometric data collected for chatbot age verification suffered a similar exposure.
NO FAKES Act: Protecting Voice and Likeness in the AI Era
Unlike the AI LEAD Act, the NO FAKES Act has real viability. Introduced by Democratic Senator Chris Coons of Delaware, it has 11 co-sponsors, six of whom are Republicans. The bill seeks to protect the voice and visual likeness of all individuals against unauthorized computerized recreations using generative artificial intelligence.
It is safe to say nobody wants photorealistic representations of themselves doing embarrassing things they never did. But the NO FAKES Act goes beyond that, holding platforms liable for hosting unauthorized digital replicas and excluding digital replicas from First Amendment protection.
Sarah Montalbano, a researcher at the Center of the American Experiment, explained how the NO FAKES Act could hurt creativity in the gaming industry. Penalties of up to 25 thousand dollars would fall hardest on small developers, hobbyists, and fan communities that create non-commercial games or mods. This would incentivize developers to preemptively restrict the variety of faces, voices, and customizable features in their products.
The Enormous Benefits of AI That Are at Stake
While lawmakers from both parties rush to name, shame, and regulate their preferred villains, they are losing sight of the big picture. The potential gains of artificial intelligence for humanity are extraordinary.
The AlphaFold system uses primary amino acid sequences to predict the three-dimensional structure of proteins, reducing prediction time from years to hours and cutting the cost of early-stage drug discovery by 30% to 70%. And it exists — in the words of Taylor Barkley, director of federal government affairs at the Abundance Institute — because researchers had the freedom to launch and iterate on imperfect models openly. The AI LEAD Act, with its imposition of strict liability on developers of AIs deemed dangerous, would have discouraged exactly the kind of experimentation that produced AlphaFold.
R-Super, an algorithm developed by researchers at Johns Hopkins University, trains AI models to segment tumors — a crucial step in cancer diagnosis and treatment — in one to two minutes, compared to the 30 minutes to one hour required by radiologists without technological assistance. The Department of Energy has already deployed AI to reduce the risk of blackouts by anticipating grid disruptions and improving load forecasting. AI increased software development speed by more than 55% in one experiment, and Anthropic’s Claude Cowork assistant is so efficient at coding that its launch and updates have triggered multiple stock market drops since its debut in January. AI has also already saved American taxpayers billions of dollars through improved fraud detection. 🔬
The Risks That Cannot Be Ignored
Not everything AI touches has been positive, of course. People using the technology carelessly have made embarrassing mistakes in law, journalism, and other fields. There have been tragedies related to AI as well. In February 2024, Sewell Setzer III, age 14, took his own life after reportedly becoming obsessed with an AI chatbot on the Character.ai platform. Fourteen months later, Adam Raine, age 16, did the same after ChatGPT reportedly provided a detailed script for ending his life, according to the lawsuit filed by his parents against OpenAI.
But no technology should be evaluated exclusively by its harms. More than 40 thousand Americans die in car accidents every year. Yet no sensible leader would propose banning motor vehicles — not only because AI will likely reduce that number by automating cars and trucks, but because the benefits of automobiles, including getting people to the hospital in emergencies, outweigh their costs. The same principle applies to artificial intelligence.
Billion-Dollar Investments and the Tug-of-War Between States and the Federal Government
Talk is cheap. Hundreds of billions of dollars in investment is not. Venture capital firms invested 259 billion dollars in AI companies in 2025 alone, and half a trillion dollars in AI capital expenditures is projected for 2026. The magnitude of these investments indicates that the expected benefits are even greater.
But AI is under threat from lawmakers at every level. Not only do some members of Congress want to pass national laws, but Congress has been unable and unwilling to preempt the growing patchwork of state laws that threatens to hamper the technology’s growth.
Not everyone in government wants to tie AI’s hands. Senator Ted Budd, chairman of the Subcommittee on Science, Manufacturing, and Competitiveness, stated that prioritizing AI advancement without subjecting it to excessive regulation is essential to maintaining the United States’ competitive edge. Senator Ted Cruz, chairman of the Commerce, Science, and Transportation Committee, called AI a new global industrial revolution that could generate opportunities to improve quality of life, create jobs, and spur economic growth.
During a subcommittee hearing, Michael Kratsios, director of the White House Office of Science and Technology Policy, advocated for applying interstate commerce principles to prevent balkanized regulation. Six months later, it remained uncertain whether the administration would manage to preempt state AI regulations.
Trump, the White House, and the Regulatory Roller Coaster
President Donald Trump himself called AI an industrial revolution, an information revolution, and a renaissance — all at the same time. One of Trump’s first actions in his second term was to revoke his predecessor’s precautionary AI framework. He also appointed AI advocates like Kratsios and David Sacks to federal positions.
Before the passage of the One Big Beautiful Bill Act in July, Republicans in Congress appeared united in wanting to shield AI from state-level strangulation. The House version included a direct 10-year moratorium on states and localities limiting, restricting, or otherwise regulating artificial intelligence. In the Senate version, Cruz proposed denying access to 42 billion dollars in broadband deployment funds to states that passed AI laws.
But then several Republicans defected from the pro-AI side. Senator Marsha Blackburn, a Republican from Tennessee, joined forces with Democratic Senator Maria Cantwell to strip the AI conditions on broadband funding from the final version of the reconciliation bill, denouncing Cruz’s proposal as a way for Big Tech to exploit children, creators, and conservatives.
Trump continued to push for a light-touch approach to AI regulation, posting on Truth Social that the U.S. should have a single federal standard instead of a patchwork of 50 state regulatory regimes. In December, he signed an executive order conditioning the release of certain broadband funds on whether or not state laws conflicted with the White House AI Action Plan — with explicit carve-outs for laws on child safety, data center infrastructure, and local government use.
The Anthropic Case and the Unexpected Turn
The bad news is that the Trump administration took an unexpected turn in its relatively liberal approach to AI in late February. Anthropic CEO Dario Amodei refused to update the terms of service for his AI model’s use by the Pentagon, insisting on maintaining explicit prohibitions on domestic mass surveillance and fully autonomous weapons systems.
In retaliation, Trump banned all federal agencies from contracting with Anthropic, and Secretary of Defense Pete Hegseth directed the Pentagon to classify the AI company as a supply chain risk. Anthropic’s 200-million-dollar contract with the Pentagon was terminated, and any company that wants to do business with the U.S. military must cut ties with the AI firm. This designation puts Anthropic in the same category as Huawei and drone manufacturer DJI. 😳
According to Dean Ball, who previously served as a senior technology adviser in the Trump administration, the federal government of the United States has become, by an extremely wide margin, the most aggressive regulator of artificial intelligence in the world.
What Is Really at Stake for the Future of AI
The good news is that most federal bills will probably fail — only the DEFIANCE Act, the NO FAKES Act, and the GUARD Act have real chances of passing. While the first two raise serious First Amendment concerns and the last one poses a grave threat to the privacy of AI users, none of them are likely to seriously harm the development and deployment of the LLMs that power the countless productive applications of artificial intelligence.
The central question that remains is whether the United States can create a regulatory framework that protects citizens from real harms without suffocating a technology that is already transforming medicine, science, energy, defense, justice, and virtually every sector of the economy. The answer to that question depends not just on lawmakers in Washington, but on the ability of society as a whole to understand that regulation and innovation do not have to be opposing forces — when done right, they can move forward together. 🚀
What becomes clear looking at this entire landscape is that the debate over AI regulation in the United States is far from over. Democrats and Republicans will keep clashing over the details, and the balance between protection and progress will be renegotiated with every new bill, every new hearing, and every new incident involving artificial intelligence. The only certainty is that ignoring the issue is no longer an option for anyone.
