According to Fast Company, Republican lawmakers, staffers, and tech industry lobbyists worked through the weekend crafting language to be added to the National Defense Authorization Act that would preempt states from passing laws regulating AI labs. The effort appears to be driven by tech industry lobbyists and insiders, with Senate Democrats having no visibility into the scope of the proposed moratorium. The language could potentially prevent states from passing any AI-related laws, including those focused on consumer protection or AI-related unemployment. Much uncertainty hangs over the fate of the state-level moratorium heading into Thanksgiving, with Democrats potentially not seeing the language until the vote on the must-pass military funding bill.
The Backroom Deal Problem
Here’s the thing about slipping controversial policy into must-pass legislation: it’s fundamentally anti-democratic. When you’ve got a small group of GOP lawmakers and tech lobbyists working behind closed doors on language that could kneecap state-level AI regulation, that should raise red flags for everyone. And the fact that Democrats might not even see the text until the vote? That’s not how transparent governance works.
I’ve seen this movie before with other industries. Tech companies love to complain about the “patchwork” of state regulations, but what they really want is no regulation at all. So they push for federal preemption that’s either toothless or doesn’t get passed, leaving a regulatory vacuum. Meanwhile, states that actually want to protect their citizens get their hands tied.
What’s Really at Stake Here
The scary part is we don’t even know how broad this language might be. Could it prevent states from regulating AI in hiring decisions? What about protecting consumers from algorithmic discrimination in housing or lending? These aren’t abstract concerns – we’re talking about real people’s lives being affected by AI systems that we know can be biased.
And let’s be honest – the federal government hasn’t exactly been speedy about AI regulation. States have been the laboratories of democracy on this front, with places like California and Colorado actually moving forward with meaningful AI governance. Preempting that progress would be a huge win for tech companies that want to operate with minimal oversight.
Broader Implications for Tech Governance
This isn’t just about AI – it’s about the ongoing battle between federal and state authority in tech regulation. We’ve seen similar fights over data privacy, net neutrality, and content moderation. The pattern is clear: when companies can’t get what they want at the federal level, they try to block states from acting too.
But here’s what worries me most: the defense bill is must-pass legislation. That means lawmakers are under enormous pressure to vote for it regardless of what riders get attached. It’s a perfect vehicle for sneaking in controversial policy that would never pass on its own merits. And when you combine that with the complete lack of transparency about what’s actually in this AI preemption language? That’s a recipe for bad policy that could haunt us for years.
