Can Congress Keep Up with AI? Sarah Beth Jansen & Austin Carson on Policy and Innovation

Show Notes

In this episode of Cyber Focus, host Frank Cilluffo moderates a timely and wide-ranging conversation on the future of AI policy and governance with Sarah Beth Jansen, a senior fellow at the McCrary Institute and longtime DC policy expert, and Austin Carson, founder of SeedAI and former legislative director for Rep. Mike McCaul. The discussion covers the Trump administration’s AI Action Plan, sector-specific regulatory approaches, and how Congress can play a constructive role without stifling innovation. Both guests emphasize the importance of local experimentation, procurement reform, and broad stakeholder engagement. With AI poised to shape everything from national security to everyday business operations, the episode underscores the urgent need to develop trustworthy, inclusive, and forward-looking frameworks that can scale with the technology.

Main Topics Covered:

• The White House’s AI Action Plan and why it marks a pivotal policy moment
• The case for sector-specific approaches to AI regulation
• The role of Congress in shaping balanced, innovation-friendly guardrails
• The importance of state-level initiatives like Utah’s AI sandbox
• Federal procurement as a lever for responsible AI adoption
• Hopes and fears around AI governance and public trust

Key Quotes:

Transcript

1
00:00:00,000 –> 00:00:01,000
Austin Carson [00:00:00]:
AI policy is everything. Policy. Every single member on the Hill could find something constructive to do involving artificial intelligence in their committee. On one hand, AI is frankly an excuse to crack back open issues that you may want to crack back open. On the other hand, there is kind of like this infinite fractal space of where AI is going to touch the economy, going to like touch scientific discovery.

2
00:00:01,000 –> 00:00:02,000
Frank Cilluffo [00:00:22]:
Welcome to CyberFocus from the McCrary Institute, where we explore the people and ideas shaping and defending our digital world. I’m your host, Frank Cilluffo, and this week I have the privilege to sit down with two of the leading thinkers around artificial intelligence. First, we’ve got Sarah Beth Jansen, who is a senior fellow at the McCrary Institute. Got to put that right up front and is also a longtime expert lobbyist in D.C. focused on IT issues, cyber privacy, surveillance, worked on the Hill, both the Senate Homeland Security Committee and Senate Judiciary. I’m not going to say where she got her JD but she got her undergraduate degree from Auburn University. So War Eagle.

3
00:00:02,000 –> 00:00:03,000
Sarah Beth Jansen [00:01:05]:
War Eagle.

4
00:00:03,000 –> 00:00:04,000
Frank Cilluffo [00:01:06]:
And we also have Austin Carson joining us today, who is the founder of Seed AI also has a storied and successful history on Capitol Hill. He served as legislative director to Congressman Mike McCall when he was at the House Homeland Security Committee as chairman and is doing some great things with his new nonprofit, which is we will hear about today. So, Sarah Beth, Austin, thank you for joining us today.

5
00:00:04,000 –> 00:00:05,000
Austin Carson [00:01:33]:
Thanks for having me.

6
00:00:05,000 –> 00:00:06,000
Frank Cilluffo [00:01:34]:
So we’re in the midst of summer here and obviously in a congressional recess, all eyes seem to be pointed and will increasingly be pointed to AI policy in the coming future. You had the administration recently released its AI Action Plan, which was the most detailed plan I’ve seen thus far. And you’ve got Congress trying to get its arms around some of these issues. So I thought I’d start with not a softball. But why now? You’ve had so many executives sounding the alarm on some of these issues, whether Elon Musk or founder of Nvidia or Sam Altman or so many others. Let’s start with you, Sarah Beth. Why now?

7
00:00:06,000 –> 00:00:07,000
Sarah Beth Jansen [00:02:21]:
Yeah, I think a lot of it has to do with kind of a stars aligning on the Hill and in the administration. I think some of that has to do with right now that there is all Republican control of the of the government. But even when we have seen that even in some ways last year and others, when there’s been the House, House AI Task force and Senator Schumer had his Insight forums, but I think now you have an alignment, stars are aligning with who is in control of the government. But I think we’ve seen with the AI action plan that having the administration involved is very important. And I think that a lot of members, Republicans across the board are kind of looking for some leadership from the administration. And while we’ve seen a lot of different pieces of legislation being introduced, I think this is going to be an impetus for Congress to move forward.

8
00:00:07,000 –> 00:00:08,000
Frank Cilluffo [00:03:12]:
Awesome, Sarah. Thank you, Austin. Thoughts?

9
00:00:08,000 –> 00:00:09,000
Austin Carson [00:03:14]:
Yeah, I’ll add a couple things. I mean, I think the first is we’ve been in kind of like a waiting pattern for this to some extent. Right. Like there have been a number of moments. Right. Like there was the chat GPT moment about 2023, right around when I or I guess 2022, right after I launched seed AI. And then of course, we had the deep seq moment last year. And I think we’re coming into a space where even some of the policy work that’s been going on has been going on since Trump won.

10
00:00:09,000 –> 00:00:10,000
Austin Carson [00:03:37]:
So like Michael Kratsios, who’s the current head of the Office of Science Technology Policy, Lynne Parker, who’s the number two there, and the executive director of the President’s Council of Advisors on Science Technology, have been thinking through and working through a lot of the stuff that’s in the action plan since. And then also the last two open source models that were released that are highly performant near frontier are Chinese. Kimi 2. Qwen 3. That’s a legitimate problem. And I think that the administration has and wisely stepped in and said, hey, we need to make sure that we have the resources in the United States, that we have the kind of intelligence pointed at this problem such that we can address it. And then the final piece is like we’re hitting the point where the models are actually kind of doing the things that they in theory were supposed to be doing the whole time. You know, on one hand we like created AI and largely people just like messed with Sydney inside of Bing’s thing. You know, it was more like a strange bizarro person people were talking to and then you’re getting some advice and then they’re multimodal, they can do visual, they can do that type of thing. But now we’re in a place where they’re a and by agentic it means that they can take more actions than just responding to your question.

11
00:00:10,000 –> 00:00:11,000
Austin Carson [00:04:40]:
The easiest version is just the search and return. In theory, anytime you’ve had a model that searched for something and came back, it was doing an agentic task. But now you have many, many steps that can happen. And if you’re looking at cyber security in particular, you know, Google’s already been using something that’s called a fuzzer. They have like going on to their old code base and they’ve repaired like 16 to 20% of it fairly autonomously and a number of companies are doing the same and all the cybersecurity companies, yada yada. But we’re going to have, for those of you that are familiar, Metasploit and that type of tool we’re about to have, you know, the hyper version of Metasploit, you know, we’re going to have that start hitting. The US has got to be prepared.

12
00:00:11,000 –> 00:00:12,000
Frank Cilluffo [00:05:18]:
And Love it. I do want to unpack. There are a lot of terms, a lot of buzzwords, a lot of thoughts being thrown around. I’m glad you kicked off with the agentic AI because that is near and dear to a lot of us, especially in the cybersecurity community, because the two are inextricably interwoven. And correct me if I’m wrong, but it seems like we’ve got two twin goals we’re trying to meet here. One, we cannot afford to lose the AI race against China. And you mentioned Deep Seek and you mentioned some of the that are coming out of China. That that should be eye opening. The second is we kind of need some guardrails here too, right? I mean Altman and others have been talking about got to do it now while it’s bridgeable, if not game over, too late, right? What is that right? Balance.

13
00:00:12,000 –> 00:00:13,000
Frank Cilluffo [00:06:05]:
And, and I think when you look at technology writ large, it’s always going to be out outpacing law and policy. But what, what are your thoughts on that? And I know that’s sort of a big question, but either of you want to jump in on that?

14
00:00:13,000 –> 00:00:14,000
Austin Carson [00:06:21]:
I mean, I can give you the. Okay, so this is a very interesting question because that is very well or is very much undefined, right? Like there are certain places where you have highly regulated industries that have been doing things like checking for adverse impact, like a number of different involved regulatory regimes that look like something you might ask for with AI. While we don’t have technology tests, so we don’t really have a good system of evaluations, we don’t really have a good system of something that you can point at as the like basis that’s been created in a bunch of different use cases outside of what the companies are, are doing. And so I think you have this deal where we need a lot more kind of like research and collaboration. I think a lot of time about the ICE acts are mentioned in the action plan. For instance, like having places where folks can get together and share best practices and start to make a better standard is really important. Two things that I thought were pretty instrumental or pretty close to essential if anything else is in the action plan. They call for the national AI Research resource to be funded and to be like a major focus going forward.

15
00:00:14,000 –> 00:00:15,000
Austin Carson [00:07:24]:
And that is a system of compute of data of different experts working together for some of these more foundational research projects all around the country. And so it’s supposed to include a number of test beds. And so we need places where different communities, different states that want to make it, their competencies can come in and say hey, we want to suddenly have like a new medical device here. We want to suddenly have something that involves mental health, we want to have whatever and we’re able to come in and collaboratively make this such that it is what people want. And I think a really easy to miss thing here is that it’s about trust, but not trust in the way where we’re gonna like make up what AI guardrails are gonna be and then hope people create the thing that we made up. But instead we needed to do a bit more of like co creation I believe. So one project very briefly we work on that’s involved in this is with Utah for instance. So the state of Utah has a regulatory sandbox, they have a regulatory promulgation function, they have an AI lab and they have an AI education, a nice little four corners deal.

16
00:00:15,000 –> 00:00:16,000
Austin Carson [00:08:24]:
But where for instance the community decided that there was a mental health application that they were very interested in, but they didn’t know how. They wanted to make the guardrails, they wanted it, they wanted to move very quickly, they wanted guardrails, didn’t know how to do it. So they ran this process through their device that they built here, came out with kind of regulation, regulatory language that has been widely acclaimed as both like enabling innovation and also preventing the worst things passed that law. And now you have as one of the only states that has kind of like a design to be functional and co created with community process for that application. So there’s just all this work that kind of still needs to be done. Love to see the Casey authorized as a place that’s focused on standards and innovation. It’s a place where folks can come together and do that work too. But it’s, it’s like all a society thing infinitely kind of like diffuse probabilistic technology requires us to decide how we engage with that.

17
00:00:16,000 –> 00:00:17,000
Frank Cilluffo [00:09:14]:
Austin, so much to unpack there that I do want to get into later. But it’s the state. What is the patch? Are we going to have a patchwork process? I want to get into the whole sausage making in a little bit. But before that, Sarah Beth, anything you want to foot stomp in terms of any elements of the AI action plan that you thought are most impactful or tee up? Maybe a general.

18
00:00:17,000 –> 00:00:18,000
Sarah Beth Jansen [00:09:36]:
I agree on sandboxes. I think, you know, we’ve seen that in the financial services sector for a long time using that type of approach. And I think before the federal government can come in and say that even these are what the standards should be or anything else that you need to have the ability for there to be input into that from a variety of different sources. But folks who are going to provide input into that need some sort of confidence that where they are doing that is not something that’s going to come back and hurt them later. As far as regulation and I think as far as what’s in the AI action plan that I thought was helpful, I think it’s been interesting to watch policymakers over the last couple of years watch AI all of a sudden move zoom to the top of their priority list when maybe we weren’t even talking about, maybe Austin probably was because he’s been doing so much work on it, but the rest of us were not talking about this. But I think what I liked in the action plan was to how much is involved in the AI system. It is not only the folks who are developing the models or deploying the models, but it’s everything that it’s going to take as far as the entire ecosystem, whether it’s our energy infrastructure for data systems or permitting reform. Just there’s a lot of different things I think that go into it and I like that aspect of it because I think that should make policymakers kind of take a step back and think about what the holistic approach is and that they need input from folks within all of those industries, from the private.

19
00:00:18,000 –> 00:00:19,000
Frank Cilluffo [00:10:53]:
Sector and is Congress. And this is an unfair question to an extent. But is it well situated to address these issues? The demographic may not be those that are using AI every day, but the flip side is is this is now prime time. It’s not a technology set of issues. So what is that right? Balance. What lessons Historically both of you have had great experiences on the Hill. What should we be learning from previous maybe we could have done a better situations.

20
00:00:19,000 –> 00:00:20,000
Sarah Beth Jansen [00:11:25]:
I do think cybersecurity is a good Example, I think that making something at least from the government side to not be overregulating, to be voluntary and to watch what companies can do within that and then give it some time and see if there needs to be more and then does there need to be something more sector specific to Austin’s talking about mental health or something in health care or the financial services industry that perhaps that’s where Congress can come in later. But after having learned and watched from the people who are the experts without kind of making I guess ill informed, perhaps regulatory decisions before knowing.

21
00:00:20,000 –> 00:00:21,000
Frank Cilluffo [00:12:00]:
Yes, yes, of course, Jen, I’d be curious and, and I, I forgot to mention Austin, that you also ran Government affairs in Nvidia. So and, and, and Jansen has a great quote that we need enough regulation to make it safe, but not so much to stymy innovation. And I’m paraphrasing there, but do we look at this as an AI issue or do we look at it sector based? You mentioned health, you could come up and you mentioned Sarah Beth, Financial services. What is the right way to unpack this? I mean, and where can Congress actually have greatest impact?

22
00:00:21,000 –> 00:00:22,000
Austin Carson [00:12:36]:
Yeah, I do think so. Two sides to this. I think the first side is, is that it’s easiest to consider and most rational to consider in concrete instances because it’s, it’s so simple to blast this up to the stratosphere and make it about saving or destroying the world like it’s very desir for us.

23
00:00:22,000 –> 00:00:23,000
Frank Cilluffo [00:12:52]:
A lot of the founders turn it to that discussion though, don’t.

24
00:00:23,000 –> 00:00:24,000
Austin Carson [00:12:55]:
But you know what I mean. And so it makes sense why that’s like the thing we gravitate towards is the future. We want to win the future promise and peril. But the moment, and I actually remember when I noticed this more constructive thing happening is that Kelly Wicker, who’s now the vice president of Tech Policy at the Meridian International center and I did an armchair with Sam Altman a couple of years ago maybe with all the members coming in between votes. And I noticed that they had all started to ask about areas of committee jurisdiction. Hey, I’m working on this, this cyber subcommittee. We’re trying to do X. How does AI involve Why? And I think when Congress starts to get there, it’s the closest to kind of their bailiwick.

25
00:00:24,000 –> 00:00:25,000
Austin Carson [00:13:31]:
I do think there’s the other side of it, of Congress needs support. Congress always needs support. And we all know the staffers are horrible, underpaid. There’s like seven staff, they all have 12 issues. And it’s all pretty complicated, you know, the world’s a little more complicated than we’d like it to be. And so providing that support to Congress and other lawmakers is something that, that’s why we set out and did this. You know, like you said, I was working at Nvidia’s Government affairs and this was pre lobbying and it was just me, which was a very strange situation for about three and I hope you.

26
00:00:25,000 –> 00:00:26,000
Frank Cilluffo [00:14:00]:
Have some stock from those.

27
00:00:26,000 –> 00:00:27,000
Austin Carson [00:14:02]:
So everybody says this, but I think it’s important to underline, which is I don’t. And it’s because I used all that stock to bootstrap my five 1C3 organization for you. I still a believer for a fact that we were going to be in this situation that we’re in. And so there’s a lot of that early stuff that I spent a lot of time working on that I couldn’t have raised money around because I’m just on the phone giving people thoughts of what they should do. But it was extraordinarily important. Do I wish I had like $30 million?

28
00:00:27,000 –> 00:00:28,000
Frank Cilluffo [00:14:28]:
Totally.

29
00:00:28,000 –> 00:00:29,000
Austin Carson [00:14:28]:
That would be awesome. You know, but it was super, super important to me. And so going forward as we get increasingly concrete, it’s going to be really important to give that increasingly concrete guidance. So it’s like one of the things that we’re looking at is like can we get together and we’re planning to get together different working groups across these places that kind of ill explored that are too higher dimensional still. It’s like is it fair use or not? That’s a very high level conversation that is not particularly actionable and it’s going to crack open old stuff. It’s like when manufacturing what do we need to do to make sure that robotics are competitive in the United States. But asking those kind of like more specific and high leverage questions of people that are outside of generic talking points and then trying to provide actionable recommendations and then the kind of secret sauce. And my personal opinion is then trying to kind of pilot those solutions out somewhere so we can give certainty.

30
00:00:29,000 –> 00:00:30,000
Frank Cilluffo [00:15:18]:
You know, I love you brought that up and this is my own messy brain thinking. But it’s almost the way the tech sector does have pilots, government also has pilots. But maybe that is the way to think about this a little bit is sector specific. So one of the fears I have is national security community. There’s an old adage, can’t march into the future backwards and win yesterday’s war because the adversary has a vote in the matter and changes their tactics based on your tactics, et cetera, et cetera. Et cetera. And we tend to be good at fixing things that blew up. That’s often too late.

31
00:00:30,000 –> 00:00:31,000
Frank Cilluffo [00:15:56]:
The question is, how do we get out in front of this where we don’t stymie exactly what it is we’re trying to preserve? And maybe it is sector specific. Maybe there is a role for every congressional committee to play in this because it is very much like cyber or homeland. It transcends all existing areas. But I’d be curious what your thinking is on that.

32
00:00:31,000 –> 00:00:32,000
Sarah Beth Jansen [00:16:19]:
So I think, I think having sector specific approaches can actually be helpful to Austin’s point about helping members and their staff be directed towards maybe a narrow solution in a particular area rather than trying to solve all the world’s problems with one type of bill or one type of regulatory approach. Maybe not necessarily to be over regulatory in each area, but think about what that area needs. I think, you know, we’ve kind of seen some of that in, in privacy with kids, online privacy and things like that. It’s been kind of hard to find this overall solution. But then we had the Take It down act pass earlier this year that was really focused on a specific type of activity online. And that was something that everyone, including the White House, could get behind. So I think then that that helped even from industry perspective, to be able to drive everyone kind of towards one area to focus on and discuss. And so I think that that would be helpful in terms of giving the staff things to work on.

33
00:00:32,000 –> 00:00:33,000
Sarah Beth Jansen [00:17:16]:
I think we both worked on the Hill and I think staffers are the same now. They’re very willing to engage and very willing to dive in. But it is helpful if industry and thought leaders can give them something to jumpstart what they’re doing.

34
00:00:33,000 –> 00:00:34,000
Frank Cilluffo [00:17:27]:
Great. And Austin, I want you to answer that question, but I also want you to sort of. So this also reminds me of the innovation regulation discussion, which comes to the very fore of some of these matters. But in addition to answering that question, is anyone doing it right? I want you to give some thought to that as well.

35
00:00:34,000 –> 00:00:35,000
Austin Carson [00:17:47]:
What a fun question to answer live.

36
00:00:35,000 –> 00:00:36,000
Frank Cilluffo [00:17:50]:
Start with the.

37
00:00:36,000 –> 00:00:37,000
Austin Carson [00:17:51]:
It’s okay, it’s okay. I joke. So, you know, there’s a shtick that I briefly thought I was going to try to make happen, but I decided not to because I was just like, I don’t even, I don’t even have a comms guy right now. But it was just like, look, AI policy is everything policy. Now, we were joking before I started that. You know, I started a year before Chat GPT and I remember talking with a pretty high up guy in one of the major tech companies. And he was like, why are you going to rifle shot at AI? I mean, it doesn’t make. I was like, what are you talking about? Every one of your products already has AI in it and so does everyone else’s.

38
00:00:37,000 –> 00:00:38,000
Austin Carson [00:18:20]:
You know, and so we come back to this thing that’s like, you know, first of all, I do think that every single member on the Hill could find something constructive to do involving artificial intelligence in their committee. On one hand, AI is frankly an excuse to find crack back open issues that you may want to crack back open. On the other hand, there is kind of like this infinite fractal space of where AI is going to touch the economy, going to like touch scientific discovery. Like one of my sticks when we’re talking about jobs, which I’m going to keep using this one because I feel it very strongly is like, you know, we’re coming out of this time in the 90s. That was like the end of history, the end of knowledge. We understand all things about science and our bodies in the universe. And then we come 20 years later, we’re like, oh, it turns out we don’t understand anything. This is terrible.

39
00:00:38,000 –> 00:00:39,000
Austin Carson [00:19:01]:
It’s terrible news. And I’m like, look, there is so much we have not discovered about the universe. Like we work in AI and science acceleration. How can artificial intelligence bring in more people and have people do more right? I’m like, if every single American. Let’s just imagine we’re in this future where nobody has to work. If every single American decided to spend the rest of their time trying to discover things about the universe, leveraging AI for science, we would not run out of stuff for like 100 years. Absolutely right. And so the human body, just the human body.

40
00:00:39,000 –> 00:00:40,000
Austin Carson [00:19:29]:
I’m like, we. I don’t want to get off in it, but I’m like, yeah, a thousand percent. And the fact that we live willingly in such ignorance while pretending like there’s nothing for people to do makes me crazy, you know, So I think you have both sides. I think you have one side where it’s like in every single area. How can we actually realize some of this potential and give people something constructive to do on the other side? It’s like, look, there is going to be an ongoing and very, very interesting conversation to have about the baseline technology and the development of the field. And there’s going to be a lot of folks that will have a reason to talk about it. And I think, you know, it’s not to say that it’s binary but it is to say that they should happen in kind of like respective and interesting ways because you can’t really understand what AI is going to do in practice until as kinetic contact with the world. In large part because the benchmarks are based on single turn.

41
00:00:40,000 –> 00:00:41,000
Austin Carson [00:20:13]:
Just like one say a thing and it says a thing back where it’s not really how you use it. So you have all this permutation and stuff. That’s why I’m like everything that would bring value to life is something that a committee member could go and put energy towards. Even like put aside the passing of legislation. Just point at the fact that your district has a transportation corridor and everybody in your district should be working on making that transportation corridor work with new autonomous technology so everybody can have jobs and the state can have, you know, a billion more gdp.

42
00:00:41,000 –> 00:00:42,000
Frank Cilluffo [00:20:42]:
And these aren’t either or propositions. Right. It seems to always be put into.

43
00:00:42,000 –> 00:00:43,000
Austin Carson [00:20:48]:
This box that you can cure cancer, you can kill everybody. So we got to do both at the same time, you know what I mean?

44
00:00:43,000 –> 00:00:44,000
Frank Cilluffo [00:20:54]:
Well said, Sarah.

45
00:00:44,000 –> 00:00:45,000
Sarah Beth Jansen [00:20:54]:
And I think that also we’ve seen it even in some of the in hearings and things that have gone on before and Austin mentioned it too. As far as a lot of all my clients are in the tech world and some folks brought in to talk about AI and cybersecurity or AI and copyright and then you get down to it and start talking to them. We’ve been using AI or machine learning in our products for over a decade and that’s almost a surprise to some members of Congress who haven’t really dug into it yet. So really connecting like hey, AI isn’t necessarily a new scary thing, but how can you member of Congress incorporate what you are already doing or what your priorities are party or whether it’s procurement reform or cyber security or anything else, or intellectual property and just have it be a part of what is already going on rather than something that’s new and completely unexplored.

46
00:00:45,000 –> 00:00:46,000
Frank Cilluffo [00:21:45]:
So it’s almost looking at the AI element pretty much everything. Everything, everywhere, all the time. I think it’s a movie, so.

47
00:00:46,000 –> 00:00:47,000
Sarah Beth Jansen [00:21:53]:
That’s right.

48
00:00:47,000 –> 00:00:48,000
Frank Cilluffo [00:21:53]:
I, I think that is an interesting way to look at it.

49
00:00:48,000 –> 00:00:49,000
Austin Carson [00:21:56]:
If I can say one thing on that real quick. It’s like you’re adding probabilistic technology to everything. And that’s what’s so interesting. We’ve reached the limit of what is observably done by one plus one equals two. And now it’s like what is most probable in this instance. And that’s what I Feel like it’d be good if we grab, like, wrapped our heads around. That’s what we’re grappling with.

50
00:00:49,000 –> 00:00:50,000
Frank Cilluffo [00:22:16]:
Austin, I want to pull the thread on. You mentioned Utah, and I’d like Sarah Beth’s thought on this as well. Should. So we’ve got interesting dynamics and flows going on here. I think that you’re seeing a lot of attention from this administration on a lot of the onus is going to be with industry, state, local, tribal, territorial. It adheres to and hems to the federalism sets of issues we all look at. The flip side is you don’t want patchworks of 50 different laws declaring the name of the day every single day, because then it ends up just being another paper exercise. Yes, it can be automated with AI, but it’s still a paper.

51
00:00:50,000 –> 00:00:51,000
Frank Cilluffo [00:23:01]:
What is that? Right. Balance. Should it. Should it be state, local? Should it be at the federal level? Should there be a preemptive set of issues? What are your thoughts on that?

52
00:00:51,000 –> 00:00:52,000
Austin Carson [00:23:14]:
I could jump in, please.

53
00:00:52,000 –> 00:00:53,000
Frank Cilluffo [00:23:15]:
Yeah.

54
00:00:53,000 –> 00:00:54,000
Austin Carson [00:23:17]:
If you want a Mullen, which is what you. So. So I think that. I think that there’s like a medium easy answer to it, right? Which is that, like, there we have. Federalism ain’t new. You know what I mean? Like, we have a decent idea of things that people do in a state and local level whenever some other aspect of it has been preempted over time. We’ve seen what happened with privacy. We’ve seen what happens with all the things in AI that move in the physical world or that involve people’s health or involve all of the, you know, all of the different areas of life in which the state governments are education, you know, all these different areas.

55
00:00:54,000 –> 00:00:55,000
Austin Carson [00:23:48]:
And so the question is more, in my view, like, what is clearly within that territory and what is clearly within a state’s kind of right to try to elevate itself in a way that is not like giving itself domain over the rest of the country. Right. So I think in Utah, what I really like is that what it’s doing is giving them space to. To create both enablement and safety around the things that they believe can be a relative efficiency and that the people in the state of Utah would really like to be able to take advantage of. Right now, there’s no federal. And I don’t think that there probably should be, unless, again, it becomes like a problem of interstate commerce. But there’s no federal, like, equal piece of legislation. Now there’s a.

56
00:00:55,000 –> 00:00:56,000
Austin Carson [00:24:29]:
There’s a thing in which that could become the model legislation that could go State to state. Right. But it’s something that allows them to experiment, something that allows them to put something into concrete. And again, this is about creating like, like, you know, like fostering the environment in the state. And again, I think what every state ultimately kind of wants is to become the model. They get the little bit of a boost from doing it first. They become the model. But when you get into this thing where like, you know, some state has an office that determines if models are, the underlying models are safe instead of it being the federal government, like, that’s.

57
00:00:56,000 –> 00:00:57,000
Austin Carson [00:25:01]:
You’re not going to have a, like 50 different forks of each model based upon which, and you’re not going to be able to geofence it. Like, that gets all pretty crazy in my view, pretty quickly. And so you have to be able to manage that part. The flip of it is, is that we are still in this kind of like figuring a bunch of stuff out phase. And so I think that the ac, or as it’s now, is the Casey and other research institutions throughout the federal government and throughout industry kind of have some work to do to decide what the federal statute should be. You know, and it’s like, we had a great, like, I want to bring up an Adam theory joke, but I know everybody else going to get it. We had it. We had a great bit talking about the moratorium and somebody was like, like, well, you can’t just replace it with nothing.

58
00:00:57,000 –> 00:00:58,000
Austin Carson [00:25:40]:
And one of the guys like, you can replace it with freedom. And I’m like, all right, first of all, I love it. And I’m going to get shirts and hats that say replace it with freedom because that kicks ass. But second of all, it’s like, you know, the, the American people have concerns, right? And each state has concerns. They also have hopes. We did a big survey on this back in the day. I didn’t release because it confused and scared me, but it was just like very, very, very interesting. And so I think that that is the easiest way to, to gauge it is, you know, what needs to be provided such that everybody else can create things and grow their communities and have hope for their kids.

59
00:00:58,000 –> 00:00:59,000
Austin Carson [00:26:12]:
And then what are things that can be done at home that give you more space to do that as a community.

60
00:00:59,000 –> 00:01:00,000
Frank Cilluffo [00:26:16]:
Very well said, Sarah.

61
00:01:00,000 –> 00:01:01,000
Sarah Beth Jansen [00:26:17]:
I think that’s how the federal government could support that if they choose to, is giving states the freedom to and the comfort to be able to explore different things and figure out what’s important to them and what they may need to do without fear of the federal government coming and saying that you can’t do that. And I think that just goes to where it has to be. Not over regulatory. There maybe just needs to be guidelines and things like that, but nothing that gets into what may be important to Utah, that’s not important to Florida and vice versa. But I do, I do think the moratorium discussion is going to come back around. I think there was too much invested in it this summer. So I think there’s more to come on it because I think those who, I think on both sides have very strong opinions on whether or not it should exist. I think that’s going to continue to be part of the discussion, but I think think it just depends on how it, whether that happens or not depends on how long Congress can, can keep going.

62
00:01:01,000 –> 00:01:02,000
Sarah Beth Jansen [00:27:09]:
I mean, Austin talked about privacy. There hasn’t been any, you know, federal privacy legislation ever. We just keep talking about it and states keep doing what they’re going to do. And that may be the case here with AI But I think it’s, it’s important to give folks the ability to explore and figure out what works for their state.

63
00:01:02,000 –> 00:01:03,000
Frank Cilluffo [00:27:26]:
And I do think there’s a happy medium there where federal government can serve as a catalyst to sort of drive because I think everyone’s waiting for someone to lead. That’s the irony, actually. That’s the wrong way to look at it from the policy standpoint. Everyone’s waiting for someone from a technology standpoint, it’s unfettered, it’s, it’s, it’s galloping, it’s way out in front and, and, and it’s just figuring out how we get our arms around that.

64
00:01:03,000 –> 00:01:04,000
Sarah Beth Jansen [00:27:53]:
I think that may be actually though why it might be important for Congress, going back to what we were talking about before, is to have a narrow, kind of a narrow focus on, on what they, within each, their community, rather than trying to boil the ocean and have one thing that every state should follow. We already have. And that’s why I think it’s been interesting to watch AI come up in individual committees because I think it can also exist within a lot of laws that we have now. We don’t have to make new laws for AI so what does it mean in the world of copyright law? What does it mean for cybersecurity? What does it mean for antitrust? And that doesn’t mean you have to go and have a new law about it. So I think that that’s, Congress can work within what we already have and see what the states do within what they have on their law.

65
00:01:04,000 –> 00:01:05,000
Frank Cilluffo [00:28:31]:
Updating existing and modernizing it. And I would think there would have to be a steering committee out of leadership from both to be able to get this done. And this is not a red or a blue issue. This is a bipartisan issue.

66
00:01:05,000 –> 00:01:06,000
Austin Carson [00:28:43]:
Right.

67
00:01:06,000 –> 00:01:07,000
Frank Cilluffo [00:28:45]:
And I’d be curious before we close here, we’re getting near the end of our time, but what’s the one thing Congress should not do?

68
00:01:07,000 –> 00:01:08,000
Austin Carson [00:28:55]:
You know, I. That was on the prep questions and I realized, I’m like, I didn’t. I didn’t think about, okay. The one thing Congress should not do.

69
00:01:08,000 –> 00:01:09,000
Frank Cilluffo [00:29:02]:
Is, you know, none of the others were. Sorry.

70
00:01:09,000 –> 00:01:10,000
Austin Carson [00:29:07]:
I was like, savage. I was gonna say the one thing Congress should not do is neglect the need for more like, research and collaboration. Like, the big thing is, is like, we have to continue to push forward, like bodies that are congressionally, like mandated or statutory or just like kind of called together by different members of Congress. And I think those should range everywhere from the very local to the national, where you can just get folks having the conversations, because I think self determination is going to be one of the more reliable things. Like, I’ve been. We’ve traveled around the country. I’ve been surprised by some of the things people do or don’t want. I think if members of Congress go back to their bread and butter, which is talk to the people they see at church, in the grocery store about this stuff, I think that we’ll have a better and easier time of making sense of this.

71
00:01:10,000 –> 00:01:11,000
Austin Carson [00:29:50]:
I think the. The other thing is to the Congress should not forget that the pace of technology is extraordinary in this moment. We have never seen capital expenditure like we have in this moment. GPUs are now for, like, sovereign reserves. Like, this is situation has become truly extraordinary. And we have to remember that and take that, like, part buried. Like, if some of the legislation that was teed up had passed before, it’s like it would have. Several things would have fallen out, right? Like as we move into more like agentic or interoperable systems, or as we move into systems that are more about math or coding, or as we get like, different scaffold.

72
00:01:11,000 –> 00:01:12,000
Austin Carson [00:30:27]:
You know, there’s a lot of different things that have emerged and happened that are actually quite both, like, I think positive from a technology trajectory standpoint, but rather neutral in terms of how it would relate to like, the desire to or to not regulate. But now it’s like, okay, well, if you just weren’t going to regulate chatbots before, and all of a sudden now the whole point is that it like, is kind of a different thing that Makes a spreadsheet for you and then a PDF file and then like, you know, like whatever goes and sends emails on your behalf that’s a notch different, you know, so.

73
00:01:12,000 –> 00:01:13,000
Frank Cilluffo [00:30:57]:
And it’s a double edged sword like most, most technology in some ways, right? Sarah Beth, anything what, what shouldn’t Congress.

74
00:01:13,000 –> 00:01:14,000
Sarah Beth Jansen [00:31:04]:
I just think they shouldn’t be over prescriptive in any legislation they do. Whether you are targeting, you know, AI and copyright or you are trying to look at procurement rules or cybersecurity just to give the chance for the technology to work and to grow. Because we all know that Congress unfortunately does not pass legislation very quickly. And I think for them to be relevant they need to not like be so specific that they just take them themselves out of the game and maybe to also focus on what the federal government should be doing. That is what they specifically have authority over, not what private companies should be doing but, but what needs to be done to modernize federal agencies and how they use AI or how they procure AI. And if they focus there, I think other things will, will come come to pass in a more natural way, but focus on what the federal government should be doing in this area.

75
00:01:14,000 –> 00:01:15,000
Frank Cilluffo [00:31:53]:
Well said. And, and this is a very broad question but biggest hope, biggest fear over the next 18 months when it comes to AI policy or AI as a technology I or its application.

76
00:01:15,000 –> 00:01:16,000
Sarah Beth Jansen [00:32:05]:
I think just from some of how some of my clients use AI. Just thinking about, I guess kind of goes back to what I just said but biggest fear would be to not think about what actually was the technology is some of it’s doing now and what it can do in the future and have unintended consequences for smaller businesses or smaller companies. I have a company who doesn’t have anything to do with ChatGPT but they’re trying to make people’s HR functions work better. Right. And they’re a smaller company and, and their biggest problem is challenging getting into the government market and the procurement system and being a small business. Right. And not being able to get what they need done. But they’re doing it to make people’s lives more efficient and better not to have some sort of scarecrow.

77
00:01:16,000 –> 00:01:17,000
Frank Cilluffo [00:32:48]:
You brought up procurement because that is a big way that the federal government can lead. Right?

78
00:01:17,000 –> 00:01:18,000
Sarah Beth Jansen [00:32:53]:
Yes.

79
00:01:18,000 –> 00:01:19,000
Frank Cilluffo [00:32:53]:
Austin, thoughts on that? Biggest hopes, biggest fears?

80
00:01:19,000 –> 00:01:20,000
Austin Carson [00:32:56]:
Well, I’ll tag on and just say the procurement thing is a great point because it is the most straightforward way to try to see if a standard can even emerge, which I think is what we’re doing out of in part the eo but I would say that the biggest hope is that we do see increasingly large number of kind of like states and groups of Americans across the country being able to develop and deploy interesting adoption, like methods of adoption or application that can really benefit the people that live there and can help drive the activity of the federal government. And my biggest fear is that AI will be developed and regulated and considered by a group of extraordinarily small number of people in the. In the bay and in D.C. and in New York instead of people around the country. And I think that we forget that this technology is made up of all of us, right? And it still takes on the characteristics however, of those that assemble it. So it’s not just that it’s like neutrally made up of all of us. It’s a cake that’s baked. And I think the fewer people we have involved, the more dangerous it is that it’s both brittle and that it’s frankly somewhat alien.

81
00:01:20,000 –> 00:01:21,000
Austin Carson [00:33:56]:
I was at the Insight Forum a couple years ago and I was sitting around in the circle and I think it was Senator Rounds or somebody was like, you know, I. I’m trying something about the, like, getting people involved or something like that. And I was just like, guys looking. It’s like, you know, Max Tech Mark and Steve Case and kind of like Aiden, like kind of an interesting group of people. It’s like, the thing is, this stuff is made up of all of our dreams. If you look around this room, our dreams are weird. You know, it was like, should not be our dreams that make up the future of the country. It should be something that’s bigger and more beautiful and frankly more American and exploratory.

82
00:01:21,000 –> 00:01:22,000
Frank Cilluffo [00:34:29]:
You know, that’s very well said. And I mean, yeah, that is the truth of it. And it’s also in its application. So there’s thousands of examples of where technology existed, but it was all in how it was applied. And this literally touches every American and everyone, period. And it’s a race we honestly cannot afford to lose. You know, I’ve asked a number of people in our podcast in Cyberfocus and almost to the T, the National Security Fund folks think AI benefits the attacker. Almost to the T, the executives and CISOs and CSOs, and they all say it benefits the defender.

83
00:01:22,000 –> 00:01:23,000
Frank Cilluffo [00:35:12]:
The reality is it benefits anyone who’s actually using it.

84
00:01:23,000 –> 00:01:24,000
Austin Carson [00:35:15]:
Right.

85
00:01:24,000 –> 00:01:25,000
Frank Cilluffo [00:35:16]:
And we need to use it in a smart kind of way. And that’s unknown at this stage. So I’d be curious, what questions didn’t I ask that I should have? I mean, this is a Broad topic. But either of you jumped off been what should I have asked to make this a more grounded discussion?

86
00:01:25,000 –> 00:01:26,000
Austin Carson [00:35:34]:
I mean, I would just say, and I’ll give you a brief answer to the thing you would, which is like, where are examples of people building up that capacity or of people doing things that are like is it better for the attacker or the defender and how are we addressing either possibility? So we work with a pro, like a project in Tulsa that’s one of the tech hubs that was set up, that’s still funded, that still like passes muster with the current EDA and their whole purpose. It’s a, a, it’s a mix of a group in Tulsa called Black Tech street. And the purpose is to build new cyber security professionals. Right. So it’s this kind of partnership with I believe like the army and Microsoft and a couple others and it’s like, okay, how do we make a new generation of cyber security professionals that can help us understand the ways in which it can benefit the defender more now or maybe the attacker, depending on if the army takes them? I don’t know, you know, but we’re in this environment where like you have to build it, you just got to build it. That’s all we can do.

87
00:01:26,000 –> 00:01:27,000
Frank Cilluffo [00:36:27]:
Sarah Beth, any thoughts there? What didn’t I ask that I should have?

88
00:01:27,000 –> 00:01:28,000
Sarah Beth Jansen [00:36:30]:
And I, I think what’s been interesting to watch in all of this is seeing how companies use and apply AI in different ways. And I think that understanding that more, I think would be helpful. You know, not, not everyone is the creator of the model or the deployer of the model if they use some aspect of AI or machine learning in whatever service that they offer. And it could be something that no one in Congress is thinking about because it, because it isn’t about, you know, cyber warfare or it isn’t about competing with China. But this company here is just trying to make their product and just trying to compete with people in their industry. So I think the, what it means practically for folks who really may not be in the limelight of these debates and how the policies will affect them.

89
00:01:28,000 –> 00:01:29,000
Frank Cilluffo [00:37:17]:
Sarah Beth, Austin. Thank you. I mean the stakes are enormous, the opportunities just as big. Thank you for providing more light than heat on a complex set of issues and ultimately making sure that our conversations grounded, informed, forward looking and quite honestly, we don’t know what the future brings. But the best way to bring it is to shape it.

90
00:01:29,000 –> 00:01:30,000
Austin Carson [00:37:41]:
We make it.

91
00:01:30,000 –> 00:01:31,000
Frank Cilluffo [00:37:42]:
Thank you both for being in this fight and for helping us think this through. So thank you.

92
00:01:31,000 –> 00:01:32,000
Sarah Beth Jansen [00:37:48]:
Thank you for hosting us.

93
00:01:32,000 –> 00:01:33,000
Austin Carson [00:37:49]:
Thank you so much for having us.

94
00:01:33,000 –> 00:01:34,000
Frank Cilluffo [00:37:50]:
Awesome. Thank you for joining us for this episode of Cyberfocus. If you liked what you heard, please consider subscribing your ratings and reviews. Help us reach more listeners. Drop us a line if you have any ideas in terms of topics, themes, or individuals you’d like for us to host. Until next time, stay safe, stay informed, and stay curious.