White House A.I. and crypto czar David Sacks discusses President Donald Trump’s executive order blocking state-level AI rules and pushing for a single national standard on FOX Business’s “Kudlow.”
DAVID SACKS, WHITE HOUSE AI CZAR: Well, Larry, good to be with you. The problem that we have right now is that we’ve got 50 different states running in 50 different directions. There’s 1,000 bills going through state legislatures right now to regulate AI, and over 100 of those measures have already passed.
You’ve got strong-willed governors for everyone from Gavin Newsom to Ron DeSantis. They’re all proposing their own regulations, and what this is going to do is create a patchwork of regulatory regimes where if you’re an AI company, and not just a big company but a little company, you could just be an innovator, a startup entrepreneur. You’ve now got to figure out how to comply with 50 different regulatory regimes, and what the president has proposed here is to have a single national or federal framework so there’s just one set of regulations to comply with, and this will make it a lot easier for our innovators.
KUDLOW: Looking at these state interferences, these content interferences, information interferences, what’s out there that they’re harping on that could really, you know, screw the pooch?
SACKS: Well, there’s a lot of different things. You know, what we’re seeing is that a lot of these state politicians are using every anxiety that the public may have about AI, and they’re trying to propose a regulation around it, even though those anxieties, they may not be proven yet. There may not be data around it.
There’s a lot of regulations being formed around horror stories. It’s very much a knee-jerk reaction, and we don’t need 1,000 bills at the state level. I mean, that just tells you how out of proportion this reaction is to the real issues of AI.
Now, as part of our federal framework, we acknowledge that there needs to be some addressing of child safety concerns. We have said that we’re going to do that, and, you know, and I think that we will put forward something there in terms of our framework. But in terms of some of these unnecessary regulations, you look at a state like Colorado.
They have outlawed a thing they call algorithmic discrimination. What that means is that if an AI model creates what they call a disparate impact on a protected group, then that is illegal, and not just the business that makes the decision is liable, but the AI developer is liable. How are they even supposed to know all the ways that their AI model is going to be used?
So that is an attempt by Colorado to insert ideology into the model. What they really want to do there is force model developers to put a DEI layer in the model to avoid a disparate impact on a protected group. That raises very serious First Amendment concerns.
We don’t like seeing blue states trying to insert their woke ideology in AI models, and we really want to try and stop that.
KUDLOW: I mean, anyway, reading through the EO, the EO is put together pretty well, I thought. You know, if you let the states run wild, you’d be breaking the Interstate Commerce Clause. They would be stopping commerce.
They would be stopping the development of this valuable technology. Startups would be dead in the water if they had to worry about 10 different state regulations.
SACKS: That’s right. Well, and here’s the thing, Larry. When you’ve got an AI model and the code is written by developers in one or several states, and then it’s trained in a data center in another state, and then the model gets prompted by a user, it’s called an inference, in another state, and then the whole service is delivered over national telecommunications infrastructure, which of those states has the right to regulate that model?
Right now, they’re all laying claim to regulating that AI model, and those regulations can be in contradiction with each other. This is a totally untenable situation for AI companies. This is why we need a single federal framework.
This is classic interstate commerce.
KUDLOW: Yes, that’s what I thought. So how are we, just in the last few moments, how are we doing, Coach? How are we doing on the AI crusade?
SACKS: Well, Larry, I think we’re doing great in the AI race. As you mentioned, we are in a global competition. China is our main competitor.
We have to win this AI race against China. It’s fundamental for our economy, for our future growth. The country that leads in AI is going to be the leading economy in the world.
We have to make sure that’s us. There’s also national security considerations. AI is going to be used to develop the algorithms that are going to power drones and things like that.
So it’s very important for both our security and our economy that we win this AI race against China. We cannot afford for our entrepreneurs to be hindered by this confusing patchwork of regulation. And if I could just say, there’s only one category where we’re behind China, and that’s in optimism.
If you look at AI optimism, roughly 83% of Chinese citizens are optimistic about AI. Only 39% of Americans are optimistic about AI. And that has a lot to do with the media.
They keep constantly drumming up these fears, which are unfounded, of things like AI job loss or loss of control. There’s no evidence to support those horror stories, but yet the media keeps drumming up these horror stories. And that’s what’s driving this knee-jerk reaction of the states to regulate AI.
But if we keep going with this, and if we don’t do what President Trump is saying that we need to do, which is create that federal framework, China is going to take the lead in this AI race.
KUDLOW: Bunch of lefty Luddites, David. That’s what you got going. Bunch of lefty Luddites.
I’ll be as polite as I can. Anyway, Mr. David Sachs, White House AI advisor, thank you. We appreciate it very much.
