States and local governments would be limited in how they can regulate artificial intelligence under a proposal currently before Congress. AI leaders say the move would ensure the US can lead in innovation, but critics say it could lead to fewer consumer protections for the fast-growing technology.
The proposal, as passed by the House of Representatives, says no state or political subdivision "may enforce any law or regulation regulating artificial intelligence models, artificial intelligence systems or automated decision systems" for 10 years. In May, the House added it to the full budget bill, which also includes the extension of the 2017 federal tax cuts and cuts to services like Medicaid and SNAP. The Senate has made some changes, namely that the moratorium would only be required for states that accept funding as part of the $42.5 billion Broadband, Equity, Access, and Deployment program.
AI developers and some lawmakers have said federal action is necessary to keep states from creating a patchwork of different rules and regulations across the US that could slow the technology's growth. The rapid growth in generative AI since OpenAI's ChatGPT exploded on the scene in late 2022 has led companies to wedge the technology in as many spaces as possible. The economic implications are significant, as the US and China race to see which country's tech will predominate, but generative AI poses privacy, transparency and other risks for consumers that lawmakers have sought to temper.
"[Congress has] not done any meaningful protective legislation for consumers in many, many years," Ben Winters, director of AI and privacy at the Consumer Federation of America, told me. "If the federal government is failing to act and then they say no one else can act, that's only benefiting the tech companies."
Efforts to limit the ability of states to regulate artificial intelligence could mean fewer consumer protections around a technology that is increasingly seeping into every aspect of American life. "There have been a lot of discussions at the state level, and I would think that it's important for us to approach this problem at multiple levels," said Anjana Susarla, a professor at Michigan State University who studies AI. "We could approach it at the national level. We can approach it at the state level, too. I think we need both."
Several states have already started regulating AI
The proposed language would bar states from enforcing any regulation, including those already on the books. The exceptions are rules and laws that make things easier for AI development and those that apply the same standards to non-AI models and systems that do similar things. These kinds of regulations are already starting to pop up. The biggest focus is not in the US, but in Europe, where the European Union has already implemented standards for AI. But states are starting to get in on the action.
Colorado passed a set of consumer protections last year, set to go into effect in 2026. California adopted more than a dozen AI-related laws last year. Other states have laws and regulations that often deal with specific issues such as deepfakes or require AI developers to publish information about their training data. At the local level, some regulations also address potential employment discrimination if AI systems are used in hiring.
"States are all over the map when it comes to what they want to regulate in AI," said Arsen Kourinian, a partner at the law firm Mayer Brown. So far in 2025, state lawmakers have introduced at least 550 proposals around AI, according to the National Conference of State Legislatures. In the House committee hearing last month, Rep. Jay Obernolte, a Republican from California, signaled a desire to get ahead of more state-level regulation. "We have a limited amount of legislative runway to be able to get that problem solved before the states get too far ahead," he said.
While some states have laws on the books, not all of them have gone into effect or seen any enforcement. That limits the potential short-term impact of a moratorium, said Cobun Zweifel-Keegan, managing director in Washington for the International Association of Privacy Professionals. "There isn't really any enforcement yet."
... continue reading