The House voted to pass the Senate version of one big beautiful bill without a moratorium that would prevent states from enforcing regulations or introducing new laws on AI systems for 10 years. Image: CRGLEN via Wikimedia Commons, CC BY-SA 4.0
When US House members voted for President Trump’s Budget Settlement Act last month, a big beautiful bill, many didn’t know that the roughly 900-page document contained a 450-word section with unusual provisions.
However, once this was discovered, the protest was swift. “A lot of people just got caught up in surprise,” said Washington Sen. Maria Cantwell, who is a member of the Senate committee on commerce, science and transportation. “When I voted for one big beautiful bill, I didn’t know about this clause,” lamented Georgia representative Marjorie Taylor Green.
The shocking moratorium sparked strong opposition, uniting unlikely groups across the political spectrum, from Republican and Democrat lawmakers to child safety advocates and civil rights groups to right-wing fire truck commentator Steve Bannon. Many have seen moratoriums imposed on some of the top AI developers as Trojan horses that violate state rights and prevent them from dealing with existing and future dangers of artificial intelligence.
Without comprehensive federal regulations on AI and other emerging information technologies, some members of Congress and several states are intervening to close the gap to prevent certain harms such as AI-generating pornography, voters and consumers, spam calls, and housing rents set by algorithms. Examples include the Take it Down Act (hosted by a Republican Senator in Texas, prohibiting “unconsensual online publication of intimate visual depictions of individuals”), the Children’s Online Safety Act (sponsored by Democrat Sen. Richard Brugnal of Connecticut, and aims to expand “internet safety.” “Protection from artificial intelligence (AI) misuse or include the voices of songwriters, performers and music industry experts.”
State lawmakers and the attorney general of both parties said the moratorium not only shattered the progress the state has made, but it also fails to protect its members over the next decade. Similarly, the 40 state attorney general wrote, ”
Opposition escalated as the Senate struggled to pass the settlement bill this week, defeating the moratorium with 99 to 1 votes on Tuesday. The House voted to pass the Senate version without a moratorium on one big beautiful bill.
For moratorium supporters, primarily large tech companies, and foreign policy think tanks, there are justifications that justify taking such extreme and clearly unpopular behavior. The state law patchwork on AI says it’s expensive and time-consuming to navigate. This creates unnecessary burdens that reduce innovation and slow progress. The risk of losing AI control over China is an even more dangerous threat. It is an “existential threat” that affects national security and even the future of the world.
“I’m not the type of person I’d quote, but Vladimir Putin says that anyone who wins will determine the direction of the world going forward,” Open’s Chief Officer, Global Affairs Officer Chris Lehaan, wrote on LinkedIn last week.
Concerns about the speed of progress and competition with China have now become the dominant view of policy and strategy for leading technologies, business groups and venture capitalists.
However, that view is being scrutinized more and more. Critics, for example, have questioned the claim that China is moving forward uninterrupted by regulations. In reality, they argue that China’s AI regulatory environment is far more restrictive than the US. The underlying assumptions are also a problem. Are the world’s greatest powers really trapped in the AI race with the aim of global domination? Should the race be a zero-sum game? Is this story real or even dangerous fiction?
How do the regulatory environments between the US and China actually compare? In his post, Lehaan of Openrai writes: “The PRC (People’s Republic of China), the top American AI competitor, is completely stopped with a small number with restrictions and vital government support.”
That’s a bold claim. But not everyone is buying it.
“It’s a complete myth,” says Gary Marcus, AI scholar and professor emeritus at NYU. “China has more regulations than we do. The reality is completely different to the point of conversation you hear from certain venture capitalists and others.”
China began developing AI regulations in 2023. These management rules cover many areas of emerging information technology, including generation AI, deepfakes, synthetic media, recommended algorithms, and search engines and social media curation algorithms. Furthermore, in China, companies developing and training AI models must submit technical details of the models to the central registry.
“I’ve been in close contact with compliance lawyers for a major Chinese high-tech company, so I’m telling you that the regulatory environment is very intensive,” says Girad Abili, an assistant professor in the Department of Cross-Border Law at Peking University and an affiliate fellow at Yale Law School’s Information Society Project.
In fact, Chinese AI developers are responsible for a lot of compliance work. “In that respect, it’s already a very different world in the way they run,” Abili says. “The US must implement very dramatic regulations, including passing the Data Privacy Act, to get close to what Chinese companies have to follow.” Public rollouts of various large language models are pending regulatory approval and are behind in China. Chinese high-tech company Baidu, Inc. Ernie Bot, an AI chatbot developed by, saw a six-month delay.
Chinese people also recognize the need to stay competitive and will change regulations to avoid compromising businesses in global competition, Abiri says. For example, initially China decided to make the generated chatbots available to the public. However, I eventually changed courses. Currently there are chatbots like Deepseek that work like Openai’s ChatGpt. The DeepSeek model is open source, but in contrast to cloud-based servers, it can run locally on a user’s computer and be available to anyone to develop new applications.
However, American tech developers argue that the delays caused by performing compliance work could lead to the US delaying China with AI development. But critics ask how important it really is. “It’s only important when there’s competition, and there’s a competition where you can win decisively, and that competitive advantage can be maintained in some way,” Abili says.
In the current situation where large language models are commodities and everyone is playing in the same playbook, it is not clear how anyone can get anything beyond the very short-lived benefits.
Imaginary race? The AI Race narrative emphasizes that by competing in other countries, we must engage in a zero-sum mindset to control our future. However, Tiffany Lee, an associate professor of law at the University of San Francisco School of Law, argues that the story is not accurate. That is, there is no basis for what society can gain from AI, as it is still unknown. “The future of artificial intelligence is not a zero-sum game.
The evidence of the AI story of weapons race is weak, and is primarily promoted in the West. Often, they are promoted by actors who stand up to make a direct profit from it. The overall premise,” says Óhéigerataigh. Framing as a high-stakes race for AI development and existential security threats to certain countries can justify extraordinary measures that bypass safety regulations, monitor surveillance and reduce legal costs for major technologies.
“All businesses wanted this, so they made it possible to save money, but the fact is that the federal government hasn’t put in a consistent AI policy, and states are trying to protect their citizens and should be able to do that,” Marcus says. “The fact that (the Senate) has hit this crazy clause is a huge victory for humanity over corporate lobbying.”
Federal regulations vacuum. The proposed moratorium on state-level regulation of AI was in a way non-action. Even those who don’t believe in state-level regulations are always suitable for borderless technologies like AI, and they still hope that the US will implement a national regulatory system.
“I’m optimistic about the possibilities of AI. In fact, no one believes we have important implications for a global race. But this moratorium proposal didn’t replace what the states are doing with a unified national framework. It replaced state-level AI regulations with nothing, he said.
Others agree. “Writing a set of guidelines that preempt the federal government’s set of guidelines on AI and what this is is a suspension of federal laws on state law without replacing it with anything,” the CEO of Tech Policy Group Chamber and former CEO of Google’s US Policy Office told ABC News.
“We’re going to prevent the state from doing anything in order to take a step to say we’re not doing anything. As far as I know, it’s unprecedented,” Larry Norden, vice president of election and government programs at the Brennan Center in New York, told NBC News. “It’s really dangerous given the stakes of this technology.”
Some observers are optimistic that Americans will not find themselves in an environment that is completely unregulated in the future, given that people already understand the high cost of not regulating social media. Also, many actors outside the US want to regulate AI. The two are the digital empires of China and the European Union. “The bottom line,” Abili said, “This is not a competition where you can win. This is basically a competition where anyone can lose if they don’t regulate this.”
In fact, the odds of federal-level regulatory frameworks could be high now that the moratorium has failed and “patchwork” issues remain. Next, “The government can say with a straight face, we don’t need state action, because here is the federal action.”
Editor’s Note: This work was produced with the support of The Future of Life Institute.
