When the House’s newly created Artificial Intelligence and Innovation Committee held its first meeting last week, instead of jumping straight into bills or worst-case scenarios, legislators did something rarer: They admitted they’re still learning.

Republican Rep. Justin Wilmeth, the chair of the new committee, set the tone, saying AI isn’t like most policy areas where issues mature over years — it’ll be a rapid-fire learning session where lawmakers have to deal with not only what happened yesterday, but what’s coming tomorrow.

That speed, more than any single application, is what makes AI hard to govern.

Arizona is only the third state to form a dedicated AI committee. That doesn’t guarantee leadership — but it does mean the state is choosing to engage early, before the rules harden elsewhere.

Today, the committee will hear its first pair of the 16 bills that have already been assigned to it — one regulating deepfake porn that depicts real people, and the other mandating state agencies identify opportunities to use AI and review or eliminate regulations that “unnecessarily restrict” AI innovation.

But before Wilmeth’s committee starts rewriting Arizona laws around AI, data centers and everything else under the committee’s purview, we sat down with Wilmeth to get a sense of what he hopes to accomplish, how he uses AI and how well he understands the complex and evolving topic he’s been put in charge of.

So today, we’ve got a Q&A (lightly edited for clarity and space) with the Legislature’s new AI czar.

Enjoy!

I ended up going to a couple of conferences over the offseason based on AI — it looked curious to me, and I was interested. I always want to learn stuff, even if I never become an expert. I like to read about something like, okay, so I get the premise, and maybe I move on. But with AI, I just kind of got hooked on it. And then I was talking with the speaker and leadership and they realized, like, this is an issue we need to have a committee for and get out in front of.

And so I credit them, because you usually don't have new committees halfway through a legislative cycle, right? And I think it's because they understood that this issue is impacting Arizonans right now, and it is something where people are concerned, maybe they're curious, maybe they're scared. There's a lot of curiosity and misunderstanding about it.

And we all grew up watching 2001 or the Terminator and AI is going to kill us, right? Skynet. I make that joke all the time, and part of me does wonder. But that's years down the road, and what we have with AI right now is it is impacting all of our lives on a daily basis, whether we realize it or not.

So I guess my long answer to that is that this is something that has sprung up as a need for society, and five years ago, nobody talked about it, but AI existed. If you use spellcheck, that's an AI. It's very Stone Age, right? But it's been around for, actually, several decades. It's just that now it's so overt and out in front of us.

I made a picture with me in my cowboy hat with the desert in the background. I had Grok make that. It's on my Twitter handle. And you can tell that it's fake, but like, if you were just looking at it quickly you’d be like, oh, Justin's out in the desert in the cowboy hat. This is like, harmless, AI fun, but it's an example of how impactful it is in our lives that we can take limitless pictures and we still find ways to improve it, to be happier with with our photos.

And those are some of the issues that we're starting to have brought to our attention to try to address. And there have been some other AI bills in the last couple of years that address porn. So if this committee does anything, I hope it lets Arizona see that we're taking the issue seriously. And I'm going to make it very bipartisan. I'm going to give everybody the opportunity to have discussions.

As with anything, if you take time to learn about something, it becomes inherently less scary because information is power. So at the end of the day, I hope this committee does that, and if somebody has an hour of free time on a Thursday at nine o'clock and wants to learn something about AI, they can watch my committee, and hopefully that helps.

I just try to be very plain spoken with them and let them understand that it's already happening, that it's already here, and you're not going to stop it, nor should we want it to.

Because it's a device that can be used for good or bad, much like a gun or a car. I can drive my car safely or I can play Grand Theft Auto with it. No law is going to stop a person from doing something bad. And people are going to use AI for bad. We can't stop that, because there are people with bad hearts.

But we can craft the rules when things happen. We can react and create laws, which I'm all about. But I want everybody to understand that this is an issue that is not going to go away, and we do have to take it seriously.

It's actually more recent, because, much like you, I love to write. So I'll journal my trips. But I won't let ChatGPT do that. And I didn't even have chat GPT on my phone until last year. Not because I was anti-AI, I just didn’t see how this is better than a Google search. Now that I use it, it is better.

So for me, it started when I started testing things. Like, I would ask it to create a form letter to the whatever of Utah, saying thank you for being on that committee or whatever. And it wrote the letter. And I'm like son of a bitch — this is good.

I wouldn't let it write my creative stuff, but form letters are all kind of dry. Things like that, things that we all hate to write, like a cover letter. If you're looking for a job, you just plug it in. This is the job. Here's my resume with my experience. Write me a 300-word cover letter, and it comes up with it. And how's that not a good thing?

So for things that are mundane, the cover letters of the world, the form letters of the world, those are things that AI can do and do very well because they're not flowery. There's nothing depthful. It's not poetry or a song, right?

Well, hell, I didn't even think about that. But I've been seeing these Facebook, Instagram reels and I’m now getting fed stuff about AI music. Some of it's fantastic.

I'm more of a rock fan, but I do like country. And there’s a version of Michael Jackson's “Beat It” in the tune of Chris Stapleton. I will send it to you if I'll find it this weekend or something. It’s like (starts singing Michael Jackson in country style), but it’s like really good man. I didn’t do it justice just there.

I grew up a fan of Weird Al Yankovic, and he did covers kind of. And he would always ask permission, but he actually never had to ask permission. He just did that to be polite.

But so in the case of AI music, I can see it being another form of entertainment where you're in on the joke. Yeah, Michael Jackson has passed away, we're never going to get a country Michael Jackson song. So it’s kind of no different than a cover band doing that.

It’s no different than what human can already do — it’s just AI doing it. I never want it to get to the point where we're being deceived that this is a totally real artist, but it's fake and we don't know it.

But when it's presented in a reasonable way, in an honest way, I think it's fine.

Oh, that's right, yeah! I’ve never used AI for constituents, I do that myself. I feel like that’s still for me to do. And look, as for an individual member doing that, that's their business. I understand we all have crazy schedules. I feel like I'm slow to reply sometimes, and I feel super guilty about it, but I wouldn't farm that out personally.

Now what I will say is I think there is a place for the initial inquiries from the public to go through AI, to filter out to where it needs to go, to save time. Because state agencies are inherently slow, and it takes a long time to get stuff done.

So a constituent writes in with their problem, and it instantly goes well, tell us more. What else is going on? And then when it reaches a point, it goes, okay, now I know who to give this to. Somebody will be in contact with you within two days.

If that's the kind of efficiencies we can get in state government, I think that's important.

Oh, I think we are heading to Skynet long-term. I've read enough books and stories where they're basically like, eventually this is where it’s going.

There's an old thing called Moore's Law. Basically, this guy named something Moore hypothesized that microchip technology will double every 18 months. I think three years from now, five years from now, 10 years from now, we'll have a totally unrecognizable society in different increments. I really believe that. And does that mean Skynet? Does that mean we've got robots in the streets? I don't think so.

But I watched a “60 minutes” story a month-ish ago, and it was like we're gonna go back and check in on this company we talked about in 2021 and it was a robotics company in Palo Alto. And they showed video of the robots then, and it was struggling like a baby to walk, and the humans had to help.

But now that same company, those things are running and they are not falling.

So I think that over the next few years, if this is a 100 point scale and the left is is, helpful Intern, and the 100% is Skynet, I think we're maybe at like, 12 or 14 now, and within four or five years, we'd be at like, double, yeah, we could be in the 70s or 80s or 90s in five years.

I think it's unstoppable. The train has left the station, so now it's just controlling it. Don't stand in front of a locomotive. You want to get on the locomotive and try to guide it.

I think that’s a good example. We need to be cognizant that we can't farm everything out to computers in an alternate reality. Because at the end of the day, humanity still matters in this mix.

You don't want, jokes aside, AI legislators. We're all elected because we represent our districts and we're all different.

I have no comment on that. (Laughs.) Look, maybe in 80 years, I don't know. But at this point, I still think there's a place for humanity, but it's about balancing that divide of, how can humans be a part of the process and have more efficient surroundings around them?

And I think that's where AI, it's a tool like anything else. It's a kitchen knife, it's a gun, it's a car, it's a pen, it's what you choose to do with it.

I remember I went to a conference in Portland, and the luncheon speaker was some lady who tells herself like an AI futurist. And I walked into the lunch and, like, kill it with fire — this is evil. This was in 2022 or 2023. And I just had this Skynet type of a mindset.

And after this lunch and what she talked about, I was an instant believer — because she said everything in such a plain spoken, easy to understand way.

And I understood that AI is several things: It is the future of humanity, but it's also the new space race and the new Cold War.

Because the Chinese have spent 30 years stealing our secrets and doing a damn good job of it. And the only reason that they can compete with us on any level, militarily or otherwise, is that they've got a lot of good spy operations.

And so I left that luncheon thinking like, oh, man, we have to protect AI from enemies that will use it for evil.

America's got to be the one that corners the market on this, because if we don't do it, other countries will get there. And even with our mistakes as a country, we're still the best option out there. Other countries will use it for inherent evil and the dystopian type world, and with American systems and freedom and democracy and supply and demand and open competition in the market. I think this is a realm that, if we corner it, it can be used to better all of the people's lives.

And so I left that with a very new understanding on how to approach this. And it was really then that I really got interested in it. And I'd buy a book on it, I'd buy a magazine, and just start to read about it casually. And now it's resulted in me running this committee.