You know, I think Congress in its wisdom saw this in the 90s, right? This is why we have the Federal Acquisition Streamlining Act, which is that the commercial… you have a much broader market around which you can amortize your R&D. In the commercial world, you can bring that stuff at a lower risk and with much greater speed to the DoD. We were able to deliver the Operation Warp Speed supply chain in two weeks during Operation Warp Speed because actually two years before that, we had built a very similar solution in oil and gas. You can't connect those prospectively. I didn't make that investment in oil and gas because I knew it would pay off when the nation needed it for COVID vaccine distribution. But really, if you're going after these hard problems, you can benefit whole-of-nation at some point in time. Every car camera and serial box that Americans bought actually subsidized our national security. So I think I would attack this systematically by thinking about what are the barriers that have meant that we have developed a defense industrial base and lost our American industrial base. I think the real issue here, to your point, we don't have an innovation problem. Innovation doesn't need capital. America's capital markets are the deepest and richest in the world. Dare I say, if you're unable to finance your idea, that probably tells you something about your idea in this country. But innovation does need customers. So shortening that OODA loop, the fiscal OODA loop of how quickly… I think we'd be better off spending half the money twice as quickly. It's really time, speed as a quality all of its own here, and that's how we drive up commercial adoption. It will pull these folks into the industrial base in a way that we really need. Yes, we need to cut the red tape, we need to get rid of some of these regulations. But I think the biggest barrier is encouraging adoption, empowering our people. So much of this — I couldn't agree more with Senator Reed's comments that technology… it's not a technology problem, it's actually a people problem, a leadership problem. You can't chop off a lot of our regulations. Something goes wrong, we come up with a new rule. We're trying to chop off one end of the distribution of all the things that can go wrong. You can't do that without making sure nothing can go right either.
Thank you, Mr. Sankar. Any changes to this approach of the nontraditional defense contractor?
Thank you. I think what we seek with nontraditional is the same power of the American economy, which is that people will take their private capital and put it at risk to build new things and offer it to the government, not at the taxpayer's expense. And if it works, that's great, and if it doesn't, no harm to the taxpayer. That's what you see with the nontraditional — they're raising private capital, they're putting their balance sheet at risk, they're delivering these innovations. If I was to contrast that to the traditional market, what the monopsonist prefers is I will pay you by the hour, I will control everything you're doing, I will own what you ultimately create, and then we are surprised that that category of traditional player isn't investing more in R&D. Well, I think literally we've gotten the industrial base that we've incentivized. So I think you know, my hope is actually we could find more ways of turning what we today view as the traditionals into nontraditional. That would be the alchemy that really powers our national security.
One other aspect, just observation. We all understand that the defense industrial base has shrunk dramatically from 20 years ago. A lot of that was through mergers and acquisitions, in some cases looking at a threatening young competitive company and buying it for reasons that might not be appropriate. How can we sort of stop that?
Well, I'm spending my time personally on that. So I think the antidote to the Last Supper, this consolidation wave that happened, is what we should call our first breakfast. You know, how do you — Palantir blazed the trail, survived the valley of death. I want to now lower the ladder and make it possible for many more new entrants to get there. How do I reduce the time it takes to get accreditation? How do I enable to field yourself not in an exercise that's not real, but in the actual warfighting needs, get more feedback and more scale as a consequence? We need a positive-sum mindset here. The big shrinking that happened during the Last Supper encourages zero-sum thinking, which we need to get out of.
Thank you, thank you very much, Mr. Sankar. In your work with the Department, what are some of the key factors that limit your company's ability to innovate?
I think really, if you think about our — when we first started the business, I thought our competition was going to be the primes, that the primes would be threatened by the innovation we were creating. But actually, the entity that was most threatened was the existing program of record. So it's our inability to tolerate heterogeneous innovation coming from a number of places. All innovation starts off as something that is heterodox. It's going to challenge the status quo. It's going to upset the apple cart. So we need to enable more flowers to bloom and recognize that innovation is fundamentally messy and chaotic, and any attempt to put process around it and make it clean destroys the innovation.
What does it do to the research that's going on in our small businesses when there's that kind of a halt on the program and we don't know how long it's going to last and we don't know whether it's going to be forever or if they're going to be able to resume what they're doing?
Well, what I can certainly speak to is the value of small business. So if we think about the American system, this is about David versus Goliath. We need the small business program to continue to encourage many more Davids to get out there, but we should be clear that we want David to get big. Where the small business program may be failing our existing entrepreneurs is it's just enough to keep them small — a class of indentured servants living as small businesses. But that's not what we aspire for them. We want the small guy to have an opportunity to become the next king. So if there were ways of continuing to evolve that program so that we were holding ourselves collectively more accountable to how many of our small businesses were able to get big, how many of them are now defining the next frontiers of what we're doing in defense innovation, I think the nation would be much better off.
Thank you very much. I certainly agree with that and hope that we can look at the next stages of the SBIR program to do that. Mr. Sankar, rapid acquisitions. I could not agree more that everything should be rapid. Speed is our greatest strength. The American entrepreneurial spirit of essentially when everything is on the line, we throw away the rulebook and we execute. Mr. D, you agree 100%.
Thank you, Mr. Chairman. Mr. Sankar, if there were a roundtable and your competitors were invited and not you, you'd have a problem with that?
Well, arguably that's what's happening today. Look, it happens. People need to get the best counsel they can. We need to move together. There are going to be lots of opportunities to keep competing. What we need to move away from is a big monolith approach where you had one chance to get involved, to actually every quarter we are adapting new technologies and there's a constant kind of reshuffling of who are the performers on the board. I do agree. If I was to add one thing on top of that, it's really bringing acquisition closer to the operators, to the warfighters. There's a way in which we divide these things up so cleanly and expect that acquisition can deliver on its own. Another way of thinking about your question on combat commanders is it's the answer to the monopsony. We have 13 COCOMs. We can introduce a lot more demand signal. We should be celebrating the heterogeneity and the needs across our COCOMs rather than having a unitary solution driven by the services that needs to be universal.
Mr. Sankar, as chair of the Senate DOGE Caucus, I couldn't agree more with your defense reformation paper where you state that small business program should not be welfare. I agree wholeheartedly. And in the past decade, 25 companies — they're notoriously known in my circles as SBIR mills — received 18% of all award dollars at DoD, amounting to about $2.3 billion dollars. That's a $92 million windfall per company in a program meant for small businesses. GAO reports that these frequent flyers have lower sales and investments and fewer resulting patents. We have a problem here. So Mr. Sankar, how can we eliminate this waste of taxpayer dollars and reorient the SBIR program to its original purpose as a source of merit-based seed funding?
I could not agree more. That's clearly an abuse of the intent here. One thing we could think about is time-limiting how long a company's eligible. It's not just about the size and staying below some sort of threshold, but look, we aspire for this small company to get big. I don't know if the right threshold is five years or 10 years, but there's some amount of time that we would expect you to have the opportunity to get big, and we're going to bet on other entrepreneurs in the future. The other part is more of a top-down: as we measure the efficacy of the SBIR programs, we should really be thinking about how many big companies were we able to create. I think that will help us have a clear head as we think about the next rounds of investments that we're going to make.
Having navigated the Pentagon's accounting and invoicing standards, regulatory requirements, terms of payment, all that — how has that affected your ability to do business with the DoD? You said you've been there a couple of decades, Mr. Sankar, so maybe in the early days as a smaller company it was much more intimidating. So if you want to go back in history a little bit, what was it like as a startup trying to do business with DoD?
It was quite complicated. I can't tell you the number of times we submitted invoices and somehow didn't fill out the right tick box somewhere, and the invoices would get kicked back. People always say you can count on the government to pay its bills. I think you can in the end, but perhaps not always on time, just given how Byzantine the process is. So it's not commercial. That's the reality of it. And we should be thinking about where the divergence from commercial standards helps the taxpayer, helps the government, and where is a vestige of how we built the system over time. It does act as a deterrent to new entrants coming in.
So for the small business folks that are out there listening, what would payment terms be like for a small business perhaps in the early days? What would be expected?
Well, everything is paid in arrears, of course. You can't structure it any other way. Maybe the payment terms are quite reasonable, net 30 or something like this. And then what's the difference between that and reality? Yeah, you could probably add a couple of months on that.
Ouch. Fighting the last war instead of talking about directed energy which costs 50 cents per shot rather than $70 million, the focus has been on missiles. And by the way, those missiles won't do anything with hypersonics. That's another technology that we were late on. This process has to start with acquiring the right things. New technologies win wars. Genghis Khan conquered the world because of the invention of the stirrup. The Battle of Agincourt was won by the longbow. World War I: the tank. World War II: the atomic weapon. I think this discussion has to start before we get to all the processes — we're going after the right products. Mr. Sankar, do you have any thoughts on that?
Yes, I do. I love the tank example in particular because it was the Royal Navy that built the tank. They were called tanks because the code name was 'tankers for the Eastern front' or something like that. I think this shows you — before the tank there was the land boat, which Churchill — this seems to be a hearing about Churchill in many ways. But the reason I think that's really important is it was a heterodox approach. If you had asked the Army, the Royal — the British Army — to think of what they were going to need to win World War I, they would have been wrong. In fact, they were wrong. They would have said more troops and deeper trenches. So we have to recognize that the innovation to fight and win the next war will come from the edges of our military, the people who are closest to those problems. It's very unlikely to come from this city. We wouldn't have had a nuclear Navy but for Admiral Rickover. And as Zumwalt said, the Navy had three enemies: the Soviet Union, the Air Force, and Hyman Rickover. So he was not widely loved, but I think we need more tolerance for the heretics, because these heretics end up being our heroes.
Well, I hope that — and I don't know how you inject creativity. I've observed the development of hypersonics, for example. The Chinese seem willing to fail. They do tests and fail. We have to have every test work, and that has dramatically, in my view, slowed down our development of some of these important technologies. Mr. Sankar, you're nodding your head. Is that correct? I mean, just like Starship, Elon learns more from Starship breaking up than he does from inherently waiting and slowing down to get the right perfect launch one time around.
The value — the rate of learning, the first derivative of learning, is our competitive weapon. It is how quickly we are adapting, not what are we capable of doing today. It's how much are we changing tomorrow. I could not agree more.
Thank you. Thank you, Mr. Chairman. Mr. Sankar, what kind of a difference would it make if we gave the combatant commands their own acquisition authority?
I think it is the single biggest difference that we can make here. You know, the Department of Defense is the only institution I know of that divides up supply and demand. The integration of supply and demand is the beating heart of any company, that consensus-driving process. The COCOMs handle the demand, real-world events. The services man, train, equip; they provide the supply. That would work if we really thought every COCOM and all needs were perfectly knowable and unitary across space and time here. But actually, all of our advantage comes from the fact that we might need slightly different things, and the signal for where that comes from is the combatant commander. So how do we give the people closest to the fight the ability to express a little bit of competitive demand signal? 90% of what you want is probably coming from the services, but that 10% gap is what's going to make or break us in that fight. How do we give them a little bit of budget, a little bit of authority, an ability to break the monopsony and introduce something like a free market where there's multiple demand signals coming? If we go back to a world we like, how did we have a world where every service was competing to build an ICBM? Well, maybe a COCOM commander should decide whether the Navy or the Air Force has the better idea and concept for their specific force employment or the emergent needs they're actually seeing. I think that competition will get us all to be better.
Seems like common sense. Why are we not doing that already?
I think having the luxury of having won the Cold War, we view that as duplication. We view that as wasteful. Why can't we just pick the right answer up front? I think our system is exquisitely designed to solve all problems that can be solved deductively, top down. We can think our way through it. But the promise of America is that there's so much messiness there, it's all inductive. Our system is very bad, poorly set up currently, to find the things you got to reason your way through. You got to experience it, roll up your sleeves, get dirty, and realize new insights as a consequence of doing that. I think we solve that by giving a little bit of strategic autonomy to the COCOM commanders to buy what they need and to build what they need.
So play that out. How would the services and the defense agencies react if they had to compete with another buyer?
Well, I think most people don't really like competition. Part of that's going to be a threat. But I think what would happen — if you get past the initial hysteria, the next step from that is, 'Okay, how do I actually change what I'm building so that the COCOM commander wants what I'm building?' That's where we're going to start to get the leverage. I can think about it: this is also the idea around competing programs and competing program managers that I saw in the FORGE Act. If we have a — what is the incentive for a program manager to adopt new commercial approaches that actually disrupt their existing program? I think today's incentive with a unitary effort is deny, deny, deny, pretend it doesn't exist, block it, versus actually 'I'm competing against another great American corridor down. I want to be the first person to adopt the disruptive technology so that I can win.'
Do you have a good example where the combatant command's lack of acquisition authority caused delays or even hurt the mission?
Well, I think you could look at the success of Project Maven, which really didn't come from the services. People love to deride OSD-level efforts as bureaucratic or not sustainable, but that innovation really came from the 18th Airborne, it came from CENTCOM, it came from EUCOM, it came from the Afghan NEO, it came from the emerging demand signal in the world, the crisis that had to be responded to, the learning that could only happen there, folded in capabilities that ultimately scaled.
I agree it's also been abused that I think it's a hard problem for the reason that you've already articulated. But I think one way that we could really buy this down is by doing more bake-offs, more things in parallel, getting more things fielded, because anyone can win a fiction writing contest — that has no correlation to your ability to perform. But when we have the satellites in space, we'll be able to tell one way or another. Maybe we'll decide actually we should have 50/50, maybe we should have multiple performers. Maybe we make bad decisions because we're evaluating you through a fiction writing contest instead of empirically in the field. I thought by the way the example one of you used a little bit ago — Elon Musk learning more from blowing up — I was at the Starship launch with President Trump, and it was very confusing for several of the business people there to hear Elon speak so positively about the booster that didn't come back and they had to put in the water, and like, 'But we learned so much.' That's a tough culture in our business and in government, but it's one we have to foster.
For years, the Army was buying Chinook helicopter engines from Honeywell, and Honeywell successfully lobbied Congress so its engines would be treated as commercial, and Honeywell wouldn't have to turn over the certified cost and pricing data. Now, Mr. Sankar, you're the CTO of Palantir, a billion-dollar tech company that contracts with DoD. Once Honeywell got the engine moved to a commercial engine, what do you think happened to the price? I'm not familiar, Senator.
It went up, not down, by 100%. And that's the problem we've got here too often. DoD is outgunned when it is negotiating with these giant defense contractors, which is exactly why it needs the cost and pricing data to avoid being ripped off. Now, Mr. Sankar, your company Palantir is looking to create a consortium with another defense tech company, Anduril, is that right? Yeah, to jointly bid for something called 'other transactions agreements' — since we have to give everything initials, OTAs — where the government also waives taxpayer protections on how to get pricing information. And I'm sure it is not your intent to team up with another organization in order to price gouge the military. So this next question should probably be easy here. DoD's Inspector General recommended requiring big contractors to alert military contracting offices when the price of a good or service goes up by 25%. It would move it up so other people know and can get eyes on it. Mr. Sankar, do you agree with the IG's recommendation?
I do agree. I think the price signal is part of the competitive market and encouraging more entrants and capital to efficiently be allocated to improve things.
Excellent. And will Palantir agree to do that voluntarily?
I would defer to my team here, but I don't think we would have any conceptual disagreement with that approach.
Okay, so can I treat that as a yes?
