When was the last time a Republican and a Democrat in Congress agreed on anything tech-related that actually mattered?

On June 4, 2026, Rep. Jay Obernolte (R-CA) and Rep. Lori Trahan (D-MA) dropped a 269-page discussion draft called the Great American AI Act. I think this is genuinely good news. Not perfect. Not finished. But good, and overdue.

The critics are already screaming. Civil liberties groups hate the state preemption clause. Some Democrats called it unworkable within hours of release. I do not buy that reaction. The alternative to this bill is not some better bill. The alternative is 50 different state laws pulling in 50 different directions while the most powerful AI systems on earth operate with zero federal accountability.

Why a patchwork of state laws is a gift to Big Tech

I remember watching California, Colorado, and New York each pass their own AI rules in the span of eighteen months and thinking: this is not regulation, this is theater. Colorado's AI Act takes effect June 30, 2026. New York has its own bill. Illinois passed one too. Every state draws the lines differently.

Obernolte and Trahan put it plainly in their joint op-ed: "Risks created by AI don't stop at state lines." That is not a talking point. That is a structural fact. The most powerful frontier models are built in one state and deployed in all fifty. A company like Anthropic or OpenAI does not build a separate model for Colorado residents.

“Policy for a technology this transformative can only be built to last if it's written by both parties.”

— Reps. Jay Obernolte and Lori Trahan, Bloomberg Law op-ed, June 4 2026

The bill is organized into four titles: Frontier AI Governance, Workforce, Cybersecurity, and Research and International Cooperation. That structure is smart policy. It does not just regulate the technology in isolation. It treats AI as a labor issue, a national security issue, and a global competitiveness issue all at once.

The accountability teeth that actually bite

Here is the part most coverage buries: the bill requires frontier AI developers to disclose model information, submit to mandatory third-party audits, and protect whistleblowers from retaliation. Those are real obligations with real teeth.

The draft also requires employers to give 60 days advance notice when AI is a substantial factor in a mass layoff. That is a worker protection most people do not know is in there. With over 100,000 tech industry job cuts already recorded in 2026, many attributed directly to AI, this provision is not hypothetical.

An Annenberg Public Policy Center survey found that 65% of Americans say the government has done too little to regulate AI, including 53% of Republicans. That is not a fringe position. That is a majority of the country, across party lines, demanding exactly what this bill attempts to deliver.

The preemption fight is real but the critics overreach

Now for the honest counterpunch. The ACLU formally opposed the bill's preemption language. Brad Carson of Americans for Responsible Innovation called the preemption provision a "generational mistake." These are serious people making a serious argument.

But here is where I part ways with them. The preemption is a three-year window with a sunset clause, not a permanent federal ceiling. It expires automatically without reauthorization. That is a negotiating tool, not a death sentence for state-level consumer protection. The Senate already voted 99 to 1 to strip a 10-year moratorium from a prior bill. Congress knows where the line is.

The bill also has real backing. House Speaker Mike Johnson (R-LA) offered early support. Six co-sponsors from both parties signed on at launch. This is not a vanity project.

What this moment actually means for the next decade

I have covered tech policy long enough to know that most bills die quietly. This one might too. Fisher Phillips predicts it is unlikely to advance before August recess. But the conversation it starts is not going away.

The bill incorporates text from multiple existing bipartisan measures, including the Future of Artificial Intelligence Innovation Act and the READ AI Models Act. That is years of groundwork being consolidated into one framework. Even if this draft never passes as written, it sets the terms of every AI governance debate that follows.

“This bipartisan framework is designed to meet the challenges posed by this rapidly advancing technology without smothering American innovation.”

— Rep. Lori Trahan (D-MA), official press release, June 4 2026

The question is not whether this bill is perfect. It is not. The question is whether the people opposing it have a better plan ready to go. They do not.

If you think a 269-page bipartisan draft with mandatory audits, whistleblower protections, and AI layoff disclosures is worse than the current nothing, I genuinely want to hear your argument. Tell me what you would do instead.