The SEAL Turned CEO: Brandon Tseng on the AI-Powered Future of War

0:00Brandon Tseng:The SEAL Teams is also a life of suffering and pain and torment. And so what I have come to, you know, I would absolutely claim like that helped me as an entrepreneur, having gone through the suffering, gone through the pain. So I'm used to it at this stage in the game. Yeah, that seems optimistic.

0:24This week on Bold Names, we're talking to Brandon Tseng. He's the co-founder of Shield AI. The startup is making both hardware and software for the next generation of warfare, one that puts autonomous drones on the front lines. That's next. From the Wall Street Journal, I'm Tim Higgins. This is Bold Names, where you'll hear from the leaders of the bold name companies featured in the Wall Street Journal. Today we ask, how is AI changing the battlefields of the future? And a quick note, we recorded this conversation before the war with Iran. Now, on to the show.

1:06Brandon, thank you so much for joining us today here on Bold Names. I'm excited. There's a lot going on with AI and defense tech. But first, I want to get into your founding story a little bit. Shield AI founded in 2015. Before that, you were a Navy SEAL, deployed twice in Afghanistan. And I'm just curious, how did that experience shape your understanding of what technology could do for warfare?

1:33Brandon Tseng:Yeah, no, first, Tim, thanks for having me. Really excited to chat today. Certainly my deployments in the Navy informed my perspective, which I've now come to appreciate having done S.H.I.E.L.D. for almost 11 years now. And look, I'll be honest. When I was in the SEAL teams, when I was operating, I wasn't really thinking about, hey, how could I do this mission better? Hey, what if I could do this? I had great – I thought at the time I had great technology. I had great assets. I had satellite imagery. I had Apache helicopters, Predator drones, F-18s, all at my disposal to accomplish my missions.

2:14Brandon Tseng:We had Palantir. And so when I was actually operating day in, day out, I was very happy with the tech stack that I was using. Now, it wasn't until I decided, hey, I'm going to leave the SEAL teams. I'm going to get out of the Navy, and I'm going to go do something else. And what else could I do? Did I start to think about ways that I could improve what I was doing on the battlefield? And I would say it was actually a much bigger macro forcing function than anything else at the end of the day.

2:50Brandon Tseng:I wanted to be a part of AI and autonomy. Grew up in the age of the Internet. Saw what that was. Asked myself what's going to be the next Internet. Decided it was going to be AI. Decided it was going to be autonomy. Asked myself what is the role of AI and autonomy in the military of 2035. This was back in 2015. decided it should be powering, commanding, maneuvering every single military asset under the sun. So we started by putting an AI pilot on a quadcopter to go inside buildings, which was a problem that I was very familiar with. I'm just curious, as having been a SEAL, how does that kind of shape your mindset as an entrepreneur?

3:28Does it help you? Are the lows as low as other entrepreneurs? Or is it just, is it easy compared to what you've been through?

3:36Brandon Tseng:No, look, entrepreneurship is a life of suffering. I tell our investors, I've stared into the abyss so many times. I tell people, look, it's awful every single day. And I think that's hard for people to understand or appreciate. It's like, oh, well, you guys are doing so well. You're a multi-billion dollar company. It doesn't make it any easier. You're solving the unsolvable problems day in, day out. And so it's awful is what I tell people. Now, to your point, the SEAL Teams is also a life of suffering and pain and torment. And so what I have come to, you know, I would absolutely claim like that helped me as an entrepreneur, having gone through the suffering, gone through the pain.

4:24Brandon Tseng:So I'm used to it at this stage in the game. Yeah, that seems optimistic. I think it's interesting. In 2015, when you would have been starting this company, AI right now is the hot, big thing, obviously, in tech. But a decade ago, autonomy for the physical world was really the big bubble at the time, especially when it came to cars. And so it would have been natural that you would have been thinking about, well, how can you deploy that in other realms? But I'm not sure that there was much enthusiasm for defense tech at that point. So I'm curious about that first product, the Nova, as you started to talk about it.

5:01It was about a large pizza size diameter drone with four propellers. What was it designed to do?

5:09Brandon Tseng:I would actually say it was like a small pizza. Small pizza. Okay. My bad. Pizza size drone. Yes. I guess the size of a pizza is in the eye of the beholder, I think. Yeah, yeah. It was a small quadcopter that would go inside buildings ahead of special operations forces, ahead of infantry forces, to map out these areas and find threats. The video feed and a map would be fed back to an operator on a tablet device or a cell phone, and they would actually see little red dots on the map of where people, threats, were inside these buildings. That was used on a number of different operations. I couldn't be more proud of the product.

5:49Brandon Tseng:Amazing product. Very, you know, to your point, like it was an interesting time in the defense tech world. The defense community, I don't think, knew what was going on yet. And interestingly, that was, you know, investors talk a lot about product market fit, where oftentimes I would claim entrepreneurs, they can't actually build a great product. The market's not the problem. In this case, we had this amazing product, net promoter score greater than 70, operating in a really awful market, which was the military small drone market at the time. Yeah, it wasn't quite clear yet what the case for this technology was, right?

6:28I mean, it seems obvious now, but I guess I'm wondering, if you think back to your days in the battlefield, what would having that Nova had been like for you? I mean, if you had that in your toolkit, how do you think you would have been using it?

6:41Brandon Tseng:Oh, yeah. Look, it's clearing buildings is the most dangerous operation that any warfighter can conduct. You're talking about gun battles at ranges of, you know, less than five meters, less than three meters in some circumstances. And so it looks really cool. I tell people, like, look, it looks really cool in the movies when you see it being done in any action movie. But terrifying in real life if you actually have to experience gunfire at super close range. And so the Nova quadcopter, special operations forces, they'd fly it in, it'd go inside the building, start mapping it out ahead. And you dramatically reduce the risk to your forces.

7:21Brandon Tseng:You reduce the risk to your partner forces. You know what's going on inside the area that you're operating, which allows you to make the best decisions possible. Coming up, defense contracts have long been dominated by big legacy companies. And Brandon says it wasn't easy to break in. Everybody knows you don't do defense. Like you don't build a business in defense tech. That was the, or in the defense field. The government's a horrible customer. And so, you know, fast forward to where we are today and it is a world of difference. How did that change? Stay with us.

8:06let's fast forward a little bit to today just kind of on where the industry's at a little bit of a struggle selling the business in the early days but now i have to imagine the mood has changed dramatically hundreds of billions of dollars i've been invested into defense tech in the last few years. I'd note that a year ago, you raised$240 million. And Bloomberg is now reporting that you're in talks to raise something like a billion dollars in funding at a$12 billion valuation. I don't know if you want to talk about that. But I'm curious how things are different now when you talk to investors from the early days to now.

8:44What is that level of enthusiasm like that you're feeling or you're hearing.

8:50Brandon Tseng:Yeah, sure. And no, I won't talk about what may or may not be happening right now regarding active fundraisers. But what I'll tell you is it's night and day difference. And so back in 2015 when we were raising, went up to Silicon Valley, went 30 for 30 no's. And every single one of those no's. That's brutal. Yeah. It hurts. And it's just a really tough thing to go through. And look, it was, hey, guys, really noble mission, but everybody knows you don't do defense. Like, you don't build a business in defense tech. That was the – or in the defense field. The government's a horrible customer. And so, you know, fast forward to where we are today, and it is a world of difference where, yes, you have investors, you know, reaching out, trying to figure out how to get involved with Shield AI.

9:38Brandon Tseng:Very interested in this space. Yeah. And one of the reasons I think that the conventional wisdom is you didn't want to get into defense tech was it's there's these huge players already that are already attached to the Pentagon. It's bureaucratic. There's very difficult kind of contracting process. Very hard to be the new guy to get into there. It was one of the issues. Right. A hundred percent. And we also kind of as a society had been fighting wars in a certain way for the last few generations. Right. Which I think brings us to the Ukraine situation, which has helped kind of prove the point that there are new ways to fight the battles.

10:16Right. Ukraine has really become more than a testing ground for cutting edge drone warfare. It has become kind of the way war is being fought and probably going to be fought into the future. How how is Shield AI involved in the in Ukraine? Yeah, sure.

10:36Brandon Tseng:Look, we have one of our actually two of our products. D bat is our aircraft. It is. Describe how big this thing is. Yeah. V bat is gosh, I want to say it's around like nine feet tall, 10 foot wingspan. Actually, I'll just look. This is this is a model of it right here. So it takes off vertically, transitions to horizontal flight, and then we'll fly very, very long distances looking for targets. It is the exact same mission that a Predator or Reaper drone does. We just do it in a smaller footprint for a fraction of the price. And so what we have been doing in Ukraine is we've been launching from the battlefield, from all these different spots.

11:19Brandon Tseng:We're very highly mobile. We fit in the back of a truck. HiveMine's on board, which is our AI pilot. HiveMine being the software that controls the physical aspect of this technology. And it's enabling it to operate while GPS and communications is jammed. And this is perhaps, you know, drone warfare gets talked a lot about in Ukraine. Certainly, there have been pieces that have talked about the electronic warfare side of things. But the electronic warfare in the Russia-Ukraine conflict is absolutely brutal. What you're saying is the Russians are using technology to block that signal. So then a drone that depends upon GPS then doesn't know where it is in the world.

11:57It can't fly around and do its stuff. Yeah. So you figured out a way around that. What is that way?

12:01Brandon Tseng:Yep. And I actually just want to add on top, it's not just drones. It is missiles. It is rockets. We have GPS guided rockets. We have GPS guided missiles. We have GPS guided bombs. And so you're talking about an arsenal that has been built on the backbone of this GPS technology that has now become largely obsolete on the battlefield because of this jamming. And so the way that Shield AI solves it is our AI pilot, which is the easiest way to think about an AI pilot is self-driving technology for unmanned systems. What Tesla is doing where they've got their self-driving car with a bunch of cameras and it's building a map of the world, that car is localizing itself using the map that it has built versus being wholly reliant on GPS.

12:51Brandon Tseng:We're doing the exact same thing. We're using the sensors on board these drones to build a map of the world, and we localize ourselves off of the map that we have built. And that is how we get through this GPS jamming problem, this communications jamming problem, where our systems can operate independent of those two technologies. And this is a big deal, right? Because a lot of U.S. drone makers have struggled in Ukraine because of this issue. It's a hard technology. Shield AI has invested in this technology for the past 10 and a half years. And because of those investments, we find ourselves alone in terms of what we are able to do on the battlefield.

13:30Brandon Tseng:So last year, actually, you know, last year, yeah, in Ukraine, and even before that in 2024, we were operating while GPS and communications is jammed. We were finding targets hundreds of kilometers away from the forward line of troops, further than the Ukrainians have ever been. and were able to target and destroy those Russian assets. So you're in Ukraine, another spot you're in. You got some attention recently. The U.S. Coast Guard has credited your drones, particularly the VBAT, for more than$1 billion in narcotic seizures last year alone. Is this how you kind of imagined your drones might be used as well?

14:08Brandon Tseng:Yeah, 100%. Look, they are, you need, and again, understand a lot of people might not necessarily be super familiar with military operations, but intelligence was driving every single one of my SEAL operations every single time. I just wanted more intel, more intel, more intel, because I could start to build target packages around that. You can never have enough intel. And so that's what our systems, the VBAT system is doing at the end of the day. it's providing an intelligence picture that allows people to make the best decisions possible. It's like, again, it's knowing what your adversary is doing all the time.

14:49Brandon Tseng:And look, there's a layered intelligence picture from satellites on down to aircraft like our VBATs, and we play a critical role in that intelligence picture. But absolutely, when I think about border security, maritime domain awareness, what we've been doing in the Caribbean Sea with the U.S. Coast Guard, 100 % like in our bailiwick in terms of what we've been doing. And also a lot of European countries, the European border security agency Frontex has been a customer of Shield AI. When we work with countries like Greece, not only do they care about their eastern border with Turkey, but they're very concerned about immigration and having maritime domain awareness coming from their southern maritime border.

15:30It is absolutely like something that Shield AI

15:34Brandon Tseng:focus is on. So last year, the Trump administration announced a major increase in defense spending to around$2 trillion. Army Secretary Dan Driscoll has talked about embracing, quote, startup culture and quote within the military. You have been in the military, you have been working to try to make the military a client for many years. Have you noticed a change in that culture in the past year? And if so, how is it different? You know, in the past, you would see portfolio acquisition executives or program executive officers. They would actually err towards the safe bet, right? No one ever got fired for hiring IBM is like the tech phrase.

16:18Brandon Tseng:No one ever got fired for hiring Lockheed Martin. At this stage in the game, it's become like, hey, why are you guys working? The question is like, okay, why is it that you're just working with a legacy prime? Is there a reason why you're not working with a company that can do it in a different way, that can solve the problem maybe in a different way? And it's not to say that like, look, I have all the respect in the world for our legacy aerospace and defense prime contractors. They make incredible products. They make incredible weapon systems. But again, their specialty has not been AI and autonomy.

16:53Brandon Tseng:It has not been software. Honestly, it hasn't been small systems. And so you see this need from the customer, from the warfighter for a different way of doing business. And that is what, you know, absolutely are feeling that change in the U.S. Army and SOCOM and the U.S. Air Force, the Navy, the Marine Corps. It is, again, nine day difference compared to where we were in 2015, 16, 17. So a decade ago, the feeling, the conventional wisdom in Silicon Valley was you don't want to have the Pentagon as a client. You clearly figured it out. What was the trick? No, no secret switch. I actually tell people like, look, working with the government is like banging your head against a concrete wall.

17:40Brandon Tseng:And so I just kind of like did that repeatedly. Now, the fun thing is like sometimes you break through that concrete wall and you're just like, look, I just did that with my head and that inspires you to keep going. There are no easy answers, period. At the end of the day, companies lose and they fail all the time. There is not an onus on the customer that the company succeeds. The onus is all on the company. You figure out what products are going to provide undeniable value to your customer, and then you figure out how to get it into the hands of your customer. Another thing that's kind of been equated to banging your head against the wall is manufacturing.

18:19Where do you build your drones?

18:21Brandon Tseng:We build them in Dallas, Texas, in a 200 ,000-square-foot facility called the Batcave. and looking at finding a new space, TBD, in terms of location for our expat, which will be an 850 ,000-square-foot manufacturing facility. State-of-the-art. What's your capacity now? How many can you crank out at this moment? Yeah, we're building about 16 VBATs per month. And we're measuring. And again, and this is the small version, right? These are 9-foot-tall aircraft with very large wingspans. Very heavy. These aren't one-way attack drones. They're not quadcopters. They are very sophisticated, state-of-the-art airplanes that we're building, right?

19:05Brandon Tseng:And there's an entire engine test regimen that goes around it. There's flight testing, just like Boeing or Airbus are flight testing your 737s. They're going through very intense quality checks before they ever make it out into the field. Same thing with these airplanes. Just ahead, Brandon describes how drones are as innovative and disruptive to modern warfare as a familiar piece of everyday technology. You can think of as our fighter jets, our aircraft carriers, as those desktops and laptops. You can think of unmanned systems as mobile phones. And they will become ubiquitous just like mobile phones.

19:44That's next.

19:58But I feel like I have heard you talk about you want to be able to provide customers millions of these things so they can have swarms of them out.

20:05Brandon Tseng:Yeah. So that's on the AI side. So we have pretty distinct products. So HiveMind, which is our AI pilot, our objective is to scale to 100 million AI pilots in customers' hands for sea, air, land, space applications, and eventually commercial applications, commercial robotic systems as well. And so we are working a number of different government programs. We also have a product that we call Hivemind Enterprise. It's an SDK, a software developer kit that we ship to customers. You can think of it as like the closest analogy I could come up with is like the cloud code of autonomy development, of AI pilot development.

20:44Brandon Tseng:It makes it easy to do. Let me unpack that a little bit first. So one part of the business is you make these drones, the hardware, and you put your software there and they go off into the world and do their thing. But there's also you're figuring out a way to use your software to allow other drone makers to use your software and have their businesses built around your software. Yeah. And so, look, at the end of the day, the customer wants an AI pilot on their unmanned system. And either Shield AI will participate in programs and build a bespoke AI pilot for that customer. We were just announced as the U.S.