Budget bill to ban states from regulating AI for 10 years
You have to wonder what's behind that ... although an inaguration photo does come to mind.
6/24 update, budget bill overall: Expect a lot of action starting 6/25. The Senate Parliamentarian is releasing opinions this week (click for current score) and the biggest ones are expected tonight and tomorrow. Sen. Majority Leader Thune (R-SD) is pushing very hard to get the bill passed yet this week in the Senate, although conflicts remain in both the Senate and now again in the House over Medicaid cuts in particular. See our most recent overall budget bill updates here, and expect more updates this week.
Remember this photo from the inauguration?
And the record $15M donations from tech bros for the inauguration alone?
It’s always interesting to see what’s buried in 1000+ page bills. I’m kind of a geek, so an odd section in the House version of the budget bill intrigued me. It proposed banning state regulation of AI for ten years.
The Senate immediately reacted negatively: states’ rights, of course. MTG even went on Fox to sound the states’ rights alarm, saying she never would have voted for the bill if she’d known that was in there.1 And more than 260 state legislators from all 50 states immediately opposed the ban.
And then … silence.
Suddenly unopposed, the Senate tweaked the AI regulation language, but a week ago it still wouldn’t pass the Byrd Rule. More tweaking followed and yesterday, the Senate parliamentarian gave the verdict that the revised ban wording can remain in the bill.
Before the budget bill, as concerns mounted about the rapid implementation of AI and potential harms, half of all U.S. states had already enacted legislation regulating AI deepfakes in political campaigns according to one tracker. And even before that, Trump wanted to quash AI “woke-ism.”
So … where did that budget bill insert come from? And why 10 years? What’s that about?
The alleged purpose is to instead pass a unified2 federal framework of AI regulations, preventing a patchwork of state regs that could hinder AI development and potentially create national security risks. That makes sense, but it would be more reassuring if Congress had any kind of track record at all on proactively or sanely regulating tech — unlike the EU, which remains substantially ahead of us on tech regulation, including AI regulation.
Why is the EU ahead? For one thing, there’s less tech inbreeding; most of the major tech bros and companies are Asia- or US-based. It’s possible it’s easier to be objective about the negative impact of tech when the owners aren’t sitting beside you at your inauguration, each in their $1M chair. And here’s more food for thought: The median age of EU Parliament members is 50 years. While only 18% of the US population is 65+, it’s 50% of the US Senate (median age 64.7).3
Remember Zuck explaining Facebook to Congress?
The problem, of course, is the timing of this imaginary “unified” framework. The 2024 election showed us how easily deepfakes can be used with voters, which is why Texas and Minnesota already ban deepfakes and 23 other states have regs on the books.
A “unified” bill to ban deepfakes before the 2026 midterms, since state bans will be outlawed if the AI piece remains in the budget bill and passes? Want to take a bet?
And I keep going back to that inauguration photo …
If any of this worries you, contact your US Senator here, and your House members here.
Right. She didn’t read the bill before voting for it. Bless her heart.
“Unified.” What country are they talking about?
Data is data, not ageism. With multiple senators in their 80s and 90s, we need to get real at some point about who is building the world younger generations have to live in.