As Senate Republicans rush to pass the Hodgepodge Tax and Expenditure Package (a big beautiful bill), the controversy centers around an extraordinary provision.
Congress is slow to pass regulations on AI, a rapidly evolving technology, and has left states to write their own laws. These state laws focus primarily on preventing certain harm, such as banning the use of deepfake technology to create unconsensual pornography, misleading voters about certain issues or candidates, or mimicking the voices of music artists without permission.
Some major players leading the US AI industry have argued that a mix of state laws would hum technology unnecessarily, especially as the US is trying to compete with China. But broad opposition, including several prominent Republican lawmakers, child safety advocates and civil rights groups, says the state is a necessary breakwater against dangerous technologies that could cause unknown harm in the next decade.
The Trump administration has made it clear it wants to loosen its reins regarding AI expansion. In his first week of office, President Donald Trump signed an executive order to ease restrictions on technology and revoke “existing AI policies and instructions that serve as barriers to American AI innovation.”
Then, in February, Vice President J.D. Vice President J.D. Vance gave a speech at the AI Summit in Paris.
However, a study from the Pew Research Center in April found that far more Americans who are not AI experts were more concerned about AI risk than potential benefits.
“Congress has shown that we can't do much in this field,” Larry Norden, vice-president of the election and government programs at the Brennan Center, a New York University nonprofit advocate for democratic issues, told NBC News.
“We're going to prevent the state from doing anything in order to say we're not doing anything. As far as I know, it's unprecedented. It's really dangerous, especially given the interests of this technology,” Norden said.
The provisions for the omnibus package were introduced by the Senate Commerce Committee, chaired by Ted Cruz, a Republican of Texas. Cruz's office has postponed comments to the committee. It said it states that under the proposed rules, it must require a substantial share of federal investment in AI.
On Friday, the senator said some provisions of one big beautiful bill law are subject to a 60-vote threshold to determine whether it can remain on the bill, but the AI moratorium is not one of them. Senate Republicans said they are aiming to vote on the bill on Saturday.
All Senate Democrats are expected to vote against the omnibus bill. However, some Republicans say they are opposed to state moratoriums passing AI laws.
Josh Hawley of Arkansas, Jerry Moran of Kansas, and Ron Johnson of Wisconsin.
Georgia Rep. Marjorie Taylor Greene, posted to X earlier this month and signed the House version of the bill, didn't realize the state wouldn't create its own AI law.
“Full transparency, I didn't know about this section,” Green wrote. “We don't know what capabilities AI will be able to demonstrate over the next decade. It's potentially dangerous to tie free reins and state hands.”
Tennessee Sen. Marsha Blackburn, a Republican on the Commerce Committee, said he is opposed to the 10-year suspension.
“We cannot ban Americans from protecting AI from harm across the country, including Tennessee's vibrant creative communities,” she said in a statement provided to NBC News. “For decades, Congress has been able to control virtual spaces and prove that it is unable to pass laws to prevent vulnerable individuals from being exploited by big technology.”
State lawmakers and the attorney generals of both parties are also opposed to AI provisions. The open letter, signed by 260 state legislators, expressed “strong opposition” to the moratorium. “Over the next decade, AI will raise some of the most important public policy issues of our time. It is important that state policymakers maintain their capacity to respond,” reads the letter.
Similarly, 40 state attorney generals of both parties opposed the provisions in letters to the legislature. “The impact of such a widespread moratorium will wipe out and completely destroy the rational state efforts to prevent known harm related to AI,” they wrote.
An analysis by the Brennan Center found that the suspension would overturn 149 existing state laws.
“State regulators are trying to enforce laws to protect their citizens, and they are instituting common sense regulations that seek to protect the worst kind of harm that is emerging on them from their members,” says Sarah Meyers West, co-op of the AI Now Institute, who calls for AI to benefit NBC.
“They say they need to wait 10 years before protecting people from AI abuse. These are alive. They are affecting people now,” she said.
Tech companies like AI and Google and Microsoft have argued that a pause is necessary to compete with China in the industry.
“There is growing awareness that the current patchwork approach to regulating AI is not working and staying in this path will continue to get worse,” Chris Lehane, Openai's Chief Global Affairs Officer, wrote on LinkedIn. “I'm not the one I usually quote, but Vladimir Putin says that anyone who wins will determine the direction of the world going forward.”
“We cannot afford to wake up to a future where 50 different states have enacted 50 conflicting approaches to AI safety and security,” Fred Humphries, corporate vice president of US government affairs at Microsoft, said in an email.
The Pro Business Lobby Chamber of Commerce has released letters signed by industry groups such as the American Independent Petroleum Association and the Meat Institute in support of the moratorium.
“Over 1,000 AI-related invoices have already been introduced at the state and local levels this year. Without a federal suspension, there will be an increase in patchwork of state and local laws that will significantly limit the development and deployment of AI,” they write.
Conversely, a diverse set of 60 civil rights groups, ranging from the American Civil Liberties Union to the digital rights groups and the NAACP, have signed their own open letters claiming that the state will pass its own AI law.
“Moratoriums can hinder the state enforcement of civil rights laws, which already prohibit discrimination by algorithms, and impact consumer protection laws by limiting the ability of both consumers and the state attorney general to seek replies on bad actors and completely eliminating consumer privacy laws,” the document says.
The Nonprofit National Centre on Sexual Exploitation opposed the moratorium on Tuesday. In particular, it highlighted how AI used minors for sexual exploitation.
AI technology generates child sex abuse material and is already being used by minors in grooming and etort, says Hailey McNamara, senior vice president of the group's strategic initiatives and programs.
“The AI moratorium on the budget bill is a Trojan horse that ends the state's efforts to curb sexual exploitation and other harm caused by artificial intelligence. This provision is highly reckless and, if passed, will lead to further weaponization of AI for sexual exploitation,” McNamara said.
