For months, the technology industry's army of Washington lobbyists has been chasing a singular, elusive goal: a comprehensive federal law that would establish a single set of rules for artificial intelligence across the entire United States. This concept, known as preemption, would override the growing patchwork of state-level AI regulations that companies argue creates confusion, increases compliance costs, and stifles innovation. But as the midterm elections approach and the political landscape shifts, this final push for preemption has taken on a new, and complicated, partner: child safety legislation.
The urgency is palpable. With Congress facing a rapidly shrinking legislative calendar before the August recess and the fall campaign season, the window for passing any major AI bill is closing. Moreover, if Democrats gain control of one or both chambers in the midterms, the pro-industry preemption approach could vanish entirely. This has forced Big Tech to make a desperate last stand, but the path forward is mired in confusion, conflicting agendas, and deep-seated skepticism.
The Preemption Dream and Its Obstacles
Preemption has long been a top priority for tech giants like Google, Meta, and OpenAI. They argue that a single federal standard would provide clarity and consistency, allowing them to develop and deploy AI systems without navigating a maze of varying state laws. States like Colorado and California have already enacted their own AI regulations, and dozens more are considering similar measures. Without preemption, the industry fears a future where every state has its own requirements for transparency, bias testing, and disclosure, creating a logistical nightmare for companies that operate nationally.
However, the push for preemption has faced significant political headwinds. On one side, conservative populists in the House are wary of any regulation that could be seen as limiting free speech or innovation. On the other, progressive Democrats demand strong protections against discrimination, privacy violations, and environmental harm. The result has been gridlock, with multiple bills introduced but none able to gather enough bipartisan support.
A Shotgun Marriage with Child Safety
In a bid to break the deadlock, the White House recently floated a controversial proposal: tie AI preemption to the Kids Online Safety Act (KOSA), a long-debated child safety bill. This legislation, originally introduced by Senator Marsha Blackburn (R-TN) and Senator Richard Blumenthal (D-CT), would require tech platforms to implement a “duty of care” to protect minors from harmful content, online predators, and other risks. By linking these two issues, the administration hopes to create a political package that appeals to both parties—offering conservatives a win on preemption while giving Democrats a victory on child safety.
But the strategy has backfired spectacularly. Reports emerged that the White House had negotiated a deal with Blackburn and child safety groups, only to leave House Republican leaders—who had passed a different, weaker version of KOSA out of committee—completely in the dark. The House version, championed by Majority Leader Steve Scalise (R-LA), significantly diluted the “duty of care” provision, a move that enraged child safety advocates. The Senate version, which passed 91-3 in 2024, is much stricter. The White House's apparent preference for Blackburn's Senate version has infuriated House Republicans, who feel blindsided and marginalized.
“No one knows really who’s driving this thing,” said a Republican lobbyist for a midsize tech company, speaking on condition of anonymity. “Everyone is deeply, deeply, deeply skeptical of the bill’s movement, because everyone is on such different pages. I think the House is not going to move anything that Blackburn wants.”
The Four Cs: A Decisive Gatekeeper
Adding another layer of complexity is Mike Davis, a Trump-allied lawyer and founder of the Article III Project. Davis previously led a successful effort to kill a Senate AI moratorium last year. He has laid out a set of conditions he calls the “Four Cs” that any preemption bill must satisfy: children, conservatives, creators, and communities. The inclusion of KOSA addresses the “children” requirement, but Davis insists all four must be met. “There is no chance in hell AI preemption will pass if it does not address the Four Cs,” he told reporters. “I will make damn sure of that. Again.”
This ultimatum puts additional pressure on negotiators, as the remaining three Cs—conservatives, creators, and communities—have different and often conflicting demands. Conservatives want assurances that the law will not infringe on free speech or impose burdensome regulations. Creators, such as artists and writers, want protections from AI-generated content that mimics their work. Communities want safeguards against discrimination and bias in AI systems. Crafting a bill that satisfies all four constituencies while also pleasing the tech industry and child safety advocates seems nearly impossible in the time remaining.
A Crowded Calendar and a Looming Deadline
Even if the political will existed, the legislative calendar is stacked with higher-priority items. Lawmakers must deal with FISA renewal, an immigration crackdown package, increased defense spending for ongoing military operations, a crypto market structure bill, affordability measures, and the deeply controversial SAVE America election reform bill. Regular budget items like Medicaid also demand attention. With only about six weeks remaining before the August recess, and then the general election campaign kicking into high gear, time is the enemy.
“It is mid June. You have a month and a half before people leave for a five-week recess. And then it’s election season,” observed an AI policy advocate. “There’s just no way.” The advocate noted that even if a bill were introduced tomorrow, it would need to go through committee markups, floor debates, and a conference committee to reconcile House and Senate versions—a process that typically takes months. A fast-track approach would require unanimous consent or a special rule, both of which are unlikely given the partisan tensions.
The potential shift in congressional control after the midterms adds another dimension. Currently, Republicans hold a slim majority in the House and a larger majority in the Senate. But polling suggests Democrats have a strong chance of flipping the House, and maybe even the Senate. If that happens, any AI legislation would likely be rewritten from scratch by Democratic leadership, who have their own priorities for AI regulation, including stronger equity and safety provisions. Why would Democrats allow a Republican-backed preemption bill to pass now, when they could draft their own version after taking power?
“After the election, what incentive do the Democrats have to support anything?” said the Republican tech lobbyist. “Like, why wouldn’t they say, ‘Fuck you, we’re gonna do our thing in the new Congress?’ I’m deeply skeptical.”
The Tech Industry’s Dilemma
The shotgun marriage of preemption and KOSA forces Big Tech into a painful choice. Do they want federal AI preemption badly enough to accept the liability exposure created by KOSA’s “duty of care” provision? For years, tech companies have fought against such duties, arguing they would lead to excessive censorship and litigation. Yet preemption, if achieved, would free them from the even more onerous prospect of complying with 50 different state regimes. Some companies may calculate that a single, albeit stringent, federal standard is better than chaos.
Others, however, are holding out for a cleaner bill that does not tie AI regulation to child safety. They worry that attaching KOSA will invite endless fights over content moderation and free speech, potentially sinking both pieces of legislation. “I can’t imagine a scenario where this bill would move,” said Austin Carson, former head of Nvidia’s government relations and founder of SeedAI, a nonprofit focused on expanding AI access. “I cannot imagine it.”
Meanwhile, child safety advocates are watching nervously. Many fear that the White House’s gambit will end up weakening KOSA if it is compromised to secure enough Republican votes for preemption. They recall the bitter fight over Section 230 reform, where similar attempts to link unrelated policies led to years of stalemate.
As the summer heats up, the fate of AI preemption—and with it, the future of AI regulation in the United States—hangs in the balance. The clock is ticking, negotiations are chaotic, and no one seems to be in charge. Whether the last desperate push will succeed or collapse under its own weight remains to be seen.
Source: The Verge News