Earlier this year, OpenAI chief executive Sam Altman appeared before Congress, underscoring the urgency of artificial intelligence policy. Since that testimony, the Capitol has become a battleground where partisan visions for AI oversight clash, even as many lawmakers prepare to leave Washington for the November election.
Contrasting positions from House leadership
On Tuesday, Republican Speaker of the House Mike Johnson (R-LA) warned that “AI is serious. We do need to make sure that it’s done safely but let’s again apply common sense.” He urged AI developers to police themselves, stating, “They don’t need the government to tell them to slow it down. If they want to slow it down, they should.” Yet Johnson added a stark caveat: “We cannot have a moratorium on the development of AI. Okay? Because then we will lose our edge to China and that has serious national security implications.”
Democratic Minority Leader Hakeem Jeffries (D-NY) took a starkly different tone on Monday, insisting that “it requires decisive congressional action immediately” to protect health, safety, and the well-being of Americans. Jeffries echoed concerns voiced by progressive figures such as Steve Bannon and Senator Bernie Sanders, who both called for stricter limits on the technology during a Washington event.
Industry’s cautious endorsement of oversight
Despite frequent disagreements, the two leading U.S. AI firms—OpenAI and Anthropic—have recently signaled openness to independent monitoring. Both companies have expressed support for watchdogs that could audit development pipelines, a stance that aligns with a bipartisan task force report released in 2024 and a joint proposal from Representatives Jay Obernolte (R-CA) and Lori Trahan (D-MA) earlier this year.
Nevertheless, the push for a cohesive national framework stalls. Republicans remain divided on the proper regulatory path, while Democrats appear content to wait for a potential majority in one or both chambers after January. The result is a legislative landscape where consensus on even the basic question—whether Congress should act at all—remains elusive.
Federal preemption vs. state-level action
The next major flashpoint involves the scope of authority between Washington and the states. Representative Kat Cammack (R-FL) warned in March that “The preemption issue is going to be one of the most difficult things that we navigate” noting that a patchwork of 50 different state regimes could cripple companies. While the GOP historically favored state-driven regulation, leaders like Rep. Steve Scalise (R-LA) have begun backing broad bans on state AI rules, a move opposed by many Democrats and libertarian-leaning Republicans.
State activists argue that red-state legislatures—particularly in Texas, Florida, and Utah—have been the most aggressive in holding AI firms accountable. Daniel Cochrane, senior fellow at the Institute for Family Studies, told NPR in June that a federal preemption could “result in inaction,” leaving critical consumer-safety protections unfunded.
Meanwhile, Center for American Progress analyst Nicole Alvarez emphasized that “the real fight is what governance looks like” stressing that even a robust federal framework must allow states to address niche concerns such as child-safety. She cited proposals to ban AI-generated child sexual abuse material as an area of potential bipartisan agreement. Senator Josh Hawley’s (R-MO) GUARD Act, for example, would criminalize sexually explicit chatbots targeting minors.
Geopolitical pressure and the race with China
Beyond domestic politics, the strategic dimension looms large. Chamber of Progress founder Adam Kovacevich warned that “we are in an AI race against China” noting that Beijing treats AI as an existential national priority. He argued that the United States must treat the competition with comparable seriousness, lest it fall behind in both innovation and security.
In sum, Congress is navigating a complex cross-currents-party rhetoric, industry self-regulation, state advocacy, and global competition-to decide how, when, and by whom AI will be governed.
