America’s fight over how far artificial intelligence should be allowed to go is now playing out on several fronts, from the White House to state capitols and Congress. President Donald Trump this week promoted a voluntary accord with major AI companies, while governors, lawmakers and some technology firms argued that stronger rules are needed.
The clash matters well beyond Washington. It will help decide who sets the limits for tools that are already showing up in public services, workplaces and online products, and whether those limits come from federal law, state law or company promises.
White House and Congress are pulling in different directions
President Donald Trump said Tuesday that top artificial intelligence companies signed a voluntary accord at the White House that would include internal and external reviews. He said he would not “stifle the growth of a technology that will be bigger than the industrial revolution,” according to the Associated Press.
Axios reported that the White House gathering produced what Trump called a “morally binding” compact on AI safety, signed by OpenAI, Anthropic, Google, Meta, xAI and Nvidia. The outlet said the accord formalizes industry self-regulation at a moment when Washington still has not passed broad AI rules.
That approach lines up with the administration’s broader policy push. In March, the White House released a national legislative framework for artificial intelligence that said Congress should not create a new federal rulemaking body for AI and warned that a “patchwork of conflicting state laws” would undermine U.S. innovation. A June executive order said the country should lead in AI by refusing to “stifle this innovation with overly burdensome regulation.”
Congress is not moving as one bloc. Sen. Josh Hawley, a Missouri Republican, and Sen. Chris Murphy, a Connecticut Democrat, announced the bipartisan AI Agent Accountability Act on Thursday. Hawley’s office said the bill would create a liability regime meant to give AI companies “every incentive to keep their products safe.”
As Washington argues, states are moving ahead. The Associated Press reported Thursday that Maryland Gov. Wes Moore is forming a bipartisan group of governors focused on AI, saying states need to respond to threats posed by the technology because of federal inaction.
California also moved this week. Gov. Gavin Newsom signed laws Wednesday aimed at protecting workers from AI risks, according to the Associated Press. AP said Newsom sharply criticized the lack of comprehensive federal AI regulation even as industry leaders warn about the technology’s dangers.
That state activity is happening despite pressure from Washington to centralize AI policy. The White House’s March framework said states should not be permitted to regulate AI development because it is “an inherently interstate phenomenon” with national security and foreign policy implications. But AP reported in June that states were still forging ahead on issues including child safety, workplace decisions and catastrophe prevention.
For residents, the result is uneven. Some protections may depend on where people live, work or go to school, while broader national standards remain unsettled. The administration has argued that a single federal policy would be more consistent, but no comprehensive federal law has passed.
The core argument is not whether AI will spread further. It is about who should decide how much risk is acceptable before powerful systems are released. Some leading AI firms are now publicly asking for more formal oversight even as the administration leans on voluntary commitments.
OpenAI said in September that it wants to work with Congress on “mandatory, capability-based national AI safety regulation.” In May, the company published a Frontier Governance Framework describing how its safety and security practices align with emerging legal requirements, including California’s Transparency in Frontier AI Act and the European Union’s AI Act code of practice.
Federal technical guidance already exists, though it is voluntary. NIST says its AI Risk Management Framework is a living document, and in July 2024 it released a generative AI profile as a companion resource. NIST says the framework is meant to help organizations manage AI risks and build trustworthiness into design, development, deployment and use.
What comes next is likely to be more conflict, not less. Governors are organizing, states are writing their own rules, some senators are proposing liability standards and the White House is still pressing self-policing over a broad new regulator. For now, the debate over how far AI should be allowed to go remains unresolved, and the answer may differ depending on whether federal officials, state leaders or the companies themselves prevail.