In Brief:
- Supporters of a ban say it would prevent a patchwork of regulations that vary by state, which could stifle AI innovation and make U.S. AI companies less competitive internationally.
- Opponents say the ban would leave Americans with no protections against the potential harms of AI, given there’s no federal law to replace state and local policies.
- State policies have tackled concerns like AI-generated child sexual abuse material, election-related deepfakes, use of algorithms to evaluate job applicants and more.
The Senate voted overwhelmingly to scrap a proposal that would’ve banned state and local governments from regulating AI for a decade.
The ban was to be included in the One Big Beautiful Bill Act. The version that passed in the House would block state and local governments from enforcing any laws or regulations specifically “limiting, restricting or otherwise regulating” automated decision systems, AI systems or AI models involved in interstate commerce — except if the laws make it easier to deploy AI.
Supporters say a ban on state and local regulation would help AI companies innovate and compete with countries like China.
House Speaker Mike Johnson has said a moratorium would prevent states from all making their own rules and stop states like California from setting the national standard or hindering innovation. AI companies like Google and Open AI supported a ban which would spare them from having to comply with requirements that could vary by state.
“We have to be careful not to have 50 different states regulating AI, because it has national security implications, right?” Johnson said recently.
But opponents of a ban point out there are no federal AI standards to replace halted state and local laws. A ban would prevent state and local governments from passing measures that would protect residents from the potential harms of AI, they argue.
In May, 40 state attorneys general argued in a letter that the ban on state and local regulation would strip away protections, leaving voters vulnerable to deepfake trickery and consumers at risk of algorithms deciding whether they get access to essential or impactful services like housing, job opportunities, health care and loans.
Organizations representing state and local governments also raised concerns in a June 26 letter.
“A 10-year pause could lead to immeasurable negative consequences without meaningful intervention from state and local governments, particularly in the likely absence of federal statutes,” wrote the heads of the National Association of State Chief Information Officers (NASCIO), the Council of State Governments, the National Association of Counties, the United States Conference of Mayors, the National League of Cities and the International City/County Management Association.
States have been pushing forward on AI policies. Already this year, all states, Puerto Rico, Washington, D.C., and the Virgin Islands introduced AI-related legislation. Also this year, 28 states as well as the Virgin Islands have adopted or enacted more than 75 new measures.
Colorado last year passed comprehensive AI regulation, while other states have focused on specific concerns like deepfakes of political candidates, AI-generated child sex abuse material and nonconsensual sexual deepfakes, and use of AI tools to evaluate job applicants.
To Ban or Not to Ban?
Sen. Ted Cruz originally introduced the proposal in the Senate as a straightforward ban on state and local AI regulation.
Lawmakers revised the approach to instead make states’ compliance with the ban a precondition for receiving federal funding that supports AI infrastructure and supports expanding broadband Internet into underserved areas.
Cruz and Republican Sen. Marsha Blackburn later considered a new compromise version of the measure that would’ve shrunk the ban to five years and included protections for children and musicians.
The death stroke came when Blackburn abandoned the compromise attempt with Cruz and instead teamed up with Democratic Sen. Maria Cantwell. Blackburn and Cantwell brought forth an amendment that would fully remove the ban from the One Big Beautiful Bill Act. This amendment drew near-unanimous support, even from Cruz. Senators voted 99-1 to drop the measure, with only Sen. Thom Tillis trying to keep it.
“The Senate came together tonight to say that we can’t just run over good state consumer protection laws,” Sen. Cantwell said in a statement. “States can fight robocalls, deepfakes and provide safe autonomous vehicle laws. This also allows us to work together nationally to provide a new federal framework on artificial intelligence that accelerates U.S. leadership in AI while still protecting consumers. “
