Two Thousand AI Bills and Still No Long-Term Plan: The Case for a National Regulator
States, Congress and the White House have flooded the zone with AI proposals. A former Nasdaq CEO argues every one fixes today's problem while leaving tomorrow's wide open.
Key points
- More than 2,000 AI-related proposals are active across U.S. Government as of 2025, according to CNBC Tech.
- Not one of them attempts to build a permanent, forward-looking federal body dedicated solely to overseeing artificial intelligence.
- Historic regulators like the SEC (Securities and Exchange Commission, the federal agency that polices financial markets) and the NRC (Nuclear Regulatory Commission, which oversees nuclear power safety) were both created only after major disasters struck.
- Bob Greifeld, former Nasdaq CEO and managing director at Cornerstone Financial Technology, calls for a dedicated national AI regulator before a crisis forces one into existence.
- Greifeld points to the SEC's public-comment model for major technology changes as a partial template worth borrowing.
More than 2,000 separate proposals to regulate artificial intelligence are circulating across American government. States account for over 1,500 of them. Congress holds hundreds more, and the White House has added dozens of executive actions on top.
That sounds like a lot of attention. It isn't.
That's the argument Bob Greifeld, former Nasdaq chief executive and now a managing director at Cornerstone Financial Technology, made in an opinion piece published by CNBC Tech. His core point: every proposal addresses a specific slice of today's AI problem, and not one lays the groundwork for a durable institution that can keep pace as the technology keeps changing.
Why does this gap matter?
Without a long-term framework, rules get written in reaction to whatever went wrong most recently, leaving the next problem uncovered until it blows up.
Greifeld draws a direct comparison to how the United States has handled disruptive technologies before. The SEC was created in 1934 after the 1929 stock-market crash wiped out millions of ordinary investors. Created after a different kind of catastrophe, the NRC came into existence following the 1979 partial meltdown at Three Mile Island, a nuclear power plant in Pennsylvania. Both agencies were born from crisis.
His argument: don't wait for the AI equivalent of a market crash or a meltdown before building the institution. It's worth noting that when we covered Australia's move to create a single national AI office on 14 July 2026, that country was acting before any comparable disaster struck, not after.
What would a national AI regulator actually do?
No formal proposal exists yet, so the shape is speculative, but Greifeld offers one concrete model to borrow from.
At Nasdaq, when engineers wanted to change the core trading system, SEC rules required the exchange to publish the details of those changes publicly and invite outside comment before anything went live. Competitors could read exactly what was coming. Greifeld found that uncomfortable at first.
Looking back, he credits that transparency requirement with helping make U.S. Capital markets the strongest in the world. He's confident that public comment on major updates to a large language model, the technology behind chatbots like ChatGPT and Claude, would attract vastly more public response than a tweak to how a stock exchange processes orders. That scale of public interest, he argues, is precisely why the model fits.
Should you worry about the patchwork that exists right now?
If you use AI tools at work or at home, the rules protecting you are uneven. Some states have passed protections; others have none. A federal body with a clear mandate would set a floor everywhere.
Greifeld is candid that a regulator will slow some innovation in the short term. He believes it's worth it: clear rules of the road, he writes, ultimately create the conditions for better outcomes for everyone.
The question is whether Washington moves before something goes badly wrong.
Common questions
Are any of the 2,000-plus AI proposals close to becoming law?
Some state-level bills have passed, covering areas like deepfakes (fake AI-generated videos or audio) and automated hiring decisions. No broad federal AI law has cleared Congress as of mid-2025.
Would a national AI regulator replace state laws?
Not automatically. Federal agencies typically set a minimum standard; states can layer additional rules on top, as they do with financial consumer protection today.
