Elon Musk’s XAI, Fiber-Optic Drones & the New Era of US Defense & Winning the AI Arms Race

Indexed: 2026-05-15

[00:00:02]Welcome everyone. We we thank you all for being here. The subcommittee will come to order. I ask unanimous consent that the chair be authorized to declare recess at any time. And without objection, so ordered. I also ask unanimous consent that non-committee subcommittee members be allowed to participate in today's hearing after all subcommittee members have had an opportunity to ask questions. Is there an objection?

[00:00:24]Without objection, not non-s subcommittee members will be recognized at the appropriate time for five minutes. Okay. Good afternoon and welcome to today's cyber information technologies and innovation subcommittee hearing on science, technology, and innovation posture at the Department of Defense. Thank you to our witnesses for being here today. We're looking forward to hearing from you and getting smarter.

[00:00:45]We live in an era of unprecedented innovation in critical areas like quantum computing, artificial intelligence, autonomy, cyber, and biotechnology. Given this progress, our nation's adversaries are working tirelessly to undermine American leadership and upset the world's security and stability. Put simply, the battlefield is changing on a fundamental level, and our opponents are dead set on using this paradigm shift and technology as an opportunity to degrade America's dominance. One of the US's greatest competitive advantages is our ability to outthink and out innovate our adversaries. At the Pentagon alone, there's an immense science and technology enterprise dedicated to advancing new ideas and technologies to arm our war fighters with necessary tools to win. Equally as important as our in-house minds is our ability to coordinate and collaborate with the other brilliant professionals at our national or our nation's academic institutions, defense industrial base and other US departments and agencies to ensure we are exploring every innovative path forward. Earlier this year, the department rolled out a massive reorganization into its research and engineering enterprise, focusing on streamlining decisions, concentrating technology investment areas, and providing clear demand signals to our industry. Four months in, I look forward to hearing about how this transition is going and the main lines of effort moving forward. And we're joined today by three DoD leaders who are critical to the advancement of American defense innovation. The honorable, excuse me, the honorable El Michael, the under secretary of defense for research and engineering, Mr. Owen West, director of the defense innovation unit, and Mr. Cameron Stanley, chief data and artificial intelligence officer.

[00:02:22]Gentlemen, thank you for being here and I look forward to your testimony. And with that, I yield to uh my friend and ranking member, Mr. Roana.

[00:02:31]>> Thank you, Mr. Chairman. Thank you for your leadership leadership. I too want to welcome each of you here for your first public hearing before this committee. I echo the chairman's comments about the rapid advance of innovation. Even having these conversations once a year seems inadequate to keeping up with the pace and breadth of technological change. and that's why your jobs are so important of the uh to that end like the chairman I'm interested to hear how the newly reconfigured research and engineering ecosystem is performing. Uh Mr. Michael you've also used the wide-ranging authorities granted as the department's chief technology officer at a greater scale than they've been leveraged before and I hope to understand how that is going. Mr. Stanley under Mr. Michael's leadership, CDAO is responsible both for maturing existing enterprise programs and the numerous tasks by the DoD's AOI strategy. I look forward to hearing your approach to prioritizing and balancing these efforts. And Mr. West, I know one of your focus areas is lowering barriers for commercial entry. That's a subject of great interest to many of us. In the interest of time, I'll stop there. I look forward to our discussion this morning. Thank you, Mr. Chairman. So gentlemen, I yield to you for the order you want to uh go and we'll yield each five minutes. Thank you.

[00:03:47]>> Uh thank you Chairman Bacon and uh thank you uh Ranking Member Connor and the distinguished members of the subcommittee. Thank you for the opportunity to testify today. I'm honored to be uh here today alongside my esteemed colleagues uh Owen West and Cameron Stanley. The three of us are here today representing a fundamental shift in how the War Department approaches innovation. We're moving from a fragmented siloed innovation ecosystem through a single unified model designed to transition techn to transition technology from the lab to the war fighter. As Secretary Heg says this this direction earlier this year, we officially realigned the innovation efforts bringing DIU, CDAO, and the strategic capabilities office under my purview as the department's chief technology officer. My mandate as the department CTO is to ensure delivery of tangible, decisive, and battlefield ready technology to our military and to do it at a velocity and scale that our adversaries cannot hope to match. The urgency stems from the hard truth that if we don't maintain technological overmatch, American service members will play pay the price in combat. The reality is heightened by the strategic environment we face today. Our adversaries are investing heavily and moving with purpose to surpass us in key technology areas that will define the f the course of future conflicts. However, the United States possesses a strategic national asset that our adversaries can never replicate. A dynamic innovation ecosystem comprised of unmatched talent and un unlimited creativity. The advantage is key disadvantage is key to our deterrence and it makes the outcome inevitable. My office remains focused on harnessing the full potential of our innovation ecosystem to serve the American war fighter. After being confirmed last May, my first order of business was to restore strategic focus on the technologies that we need to maate maintain dominance on the battlefield. I inherited a list of 14 CTAs or critical technology areas which really didn't prioritize anything. They were a laundry list. They diluted our resources and slowed progress. through rigorous datadriven analysis and in consultation with the military services and the and our experts in the intelligence community and made the decision to pair that list down to six.

[00:06:06]The new list of six CTAs provides clarity to industry and to our innovators in our labs and across the war department focusing their efforts and their funding vehicles. The War Department's new critical technology areas are applied artificial intelligence, biommanufacturing, contested logistics technologies, quantum and battlefield information dominance, scale directed energy and scaled hypersonics. The department is focused on its investments, its talent and its leadership around these areas while the priv private sector crowds the its capital around to supercharge progress in these areas. And let me be clear, other technology areas such as space and micro electronics remain important. R&D will continue to carry out our statutory mandate to provide rigorous oversight of all departmental R&D. But these six areas are the ones that will benefit most from attention, advocacy, and resources at my level. for each of the CTAs. Another thing we've changed is we've made uh tangible focused sprints that we can measure our progress against them to accelerate the technology development cycles and we don't get stuck doing nothing. Each sprint has an accountable leader and they're mandated to deliver with strict timelines and accountability. An example of this is the rollout of genai.mill which is the war department's official secure generative AI platform. We've unleashed frontier AI to 3 million uniform personnel, contractors, and civilians. In just 5 months, the platform has attracted 1.4 million of the 3 million uh people we have in the department. Uh they've generated over 69 million prompts. They've built 100,000 agents. Prior to that, we'd only had 80,000 users uh using AI in the whole department. But we're not limited limiting deployment to our unclassified systems alone. Along in the last few weeks, we signed agreements to with virtually all leading American frontier AI companies and AI infrastructure companies to deploy their capabilities onto our classified systems.

[00:08:12]We will continue to operate at a speed and scale that ensures AI is readily available to our war fighters. Our efforts have not been limited to CTAs and just AI. We've also bring uh bringing innovation to the production and acquisition of conventional weapons enabling the return of mass effects through employment of large numbers of less exquisite weapon systems. As you've seen uh yesterday, we announced the signing of five contracts with new entrance to build lowcost munitions uh for the first time in the department's history. These uh represent a new era where we can have mass traitable affordable weapons and we're not doing it on a cost plus basis. We're doing it on a fixed price basis with with delivery dates that begin in 2027.

[00:09:01]Uh lastly on the structure as uh both the chairman and ranking member have mentioned uh it allows us to lay out a clear and rapid pathway for new and innovative ideas to make its way to work from working prototype to scaling the capability for the joint force. This is directly going after the infamous valley of death. You could start at the beginning with or the reauthorization of cers sitter and then which by the way Andrew Castellian started at cers sitter program. So, it's a good note that that program has produced results. And then we get to DIU for prototypes, appetit for scaling, and the office of strategic capital to go big. If you help wrap it up, I'm worried about votes coming up in a little bit. So, got >> with that, I will pass it off to my uh partners.

[00:09:48]>> Thank you very much.

[00:09:50]>> Good or good afternoon, Chairman Bacon, Ranking Miner K, Ranking Member Kana, and distinguished members of the subcommittee. Thank you for the opportunity to discuss how the Chief Digital and AI office at the Department of War is arming our nation's most operationally demanding warf fighting organizations from tactical special operations formations to global combatant commands with AI enabled capabilities that redefine the modern battlefield and secure US military dominance for years to come. Since February, CDAO's flagship capability teams, Maven Smart System and War Data Platform, have reimagined or have remained on call around the clock in support of Operation Epic Fury, enabling US forces to strike more than 13,000 targets in 38 days. What once days now takes seconds? What enabled this shift? Historically, the department's approach to technology development siloed war fighters from technology delivery teams, producing a capability cycle where solutions were overbudget, late to need, and not fit for purpose. At the height of operations in Iraq and Afghanistan, the typ a typical intelligence analyst was required to spend 16-hour days reviewing UAV footage for tiny anomalies that might signal major threats. The scale of unstructured data combined with human exhaustion made UAV image analysis one of the clearest operational pain points and a top priority for a solution that brought humans back into the loop in a framework that no longer was manageable with humans alone. In response in two 2017 the al algorithmic warfare cross functional team also known as project maven was established to apply data and machine learning to the problem. In theory utilizing worldclass computer vision models to detect objects of interest seems seems straightforward. In practice teams encountered fragmented frameworks static slides and an operating environment that was most effective where the most effective linkage between systems was a person is in a swivel chair. These were not simply data and scale or scale problems. They were human process and information problems that even the most advanced AI systems could not solve. The Project Maven team realized that real progress requires more than better technology. It required better processes and better teams achieved by embedding engineers with war fighters to build tools that enhance operational decision-making and human judgment. Over the past nine years, that approach has culminated into Operation Epic Fury's AI enabled datacentric warfighting system in a real world conflict. But Operation Epic Fury is one operation. What does it look like to scale the Maven model across combatant commands and mission areas?

[00:12:21]The department's answer is it is its AI acceleration strategy released on January 12th. The strategy codifies seven years of operational experience and lessons into a clear executable plan. Today I will focus on a few of those critical lessons. The first lesson, the department must leverage, not compete with commercial innovation.

[00:12:41]In 2025 alone, Frontier AI labs raised over 120 billion and the industry capital expenditure on AI infrastructure exceeded 400 billion. The department should not outspend our commercial partners. We should harness their technological and economic dynamism for national defense. This approach reshapes how the DO contracts and recruits. For example, CDAO has reduced timelines for awarding contracts from years to as little as six days, bringing top engineering talent into mission workflows at speed and scale.

[00:13:10]Critically, sorry. Second lesson, the United States military is best equipped and trained fighting force in the world, not just because of what it builds, but because of its ability to field and sustain capability at the tactical edge under the most austere conditions. To meet their needs, CDAO embeds with war fighters to enable, iterate, and deliver operationally relevant capabilities. We work on their terms in their timelines so they can fight, win, and come home.

[00:13:40]Next lesson. Speed matters, especially in AI. Traditional development programs take years to produce results. Instead, CDIO is executing pace setting projects, rapid strategically scoped initiatives that address urgent adoption challenges using the maven model of engineer warfighter teams. So far, CDA has launched seven pace setting projects that are advancing are driving advances in human machine teaming, real-time operational learning and data integration in support of real world mission.

[00:14:09]Fourth lesson, today's battlefield requires modernizing the underlying software architecture. To that end, CDAO is transforming Advana into the WAR data platform and elevating it along with the Maven Smart System into departmentwide programs. WAR data platform serves as a single source of truth for operational formations while the Maven smart system will centralize its role within the department's Cad2 uh initiative and accelerate AI enabled applications across the force. In every formation, in every geography, on every mission, the American war fighter operates on CDAO technology.

[00:14:44]The last lesson, the department must win the AI infrastructure arms race. China has invested upwards of $300 billion in energy and comput infrastructure to develop sovereign AI. Moreover, in the early days of Operation Epic Fury, the world witnessed deliberate strikes from Iranian drones and missiles targeting commercial data centers in the UAE and Bahrain. They must not they will not win this arms race. As we speak, the department is aligning with its uh partners in the department of energy, intelligence community, and inter agency to build the AI arsenal, a modular ecosystem with a three- tiered strategy to ensure compute power outmasters our joint capacity needs. It will always be.

[00:15:23]Yes, Mr. Chairman. Mr. Chairman, members of the subcommittee, the modern battlefield is evolving at an pace unmatched in recent history. In an era of heightened geopolitical competition, our ability to achieve decision advantage is essential to maintaining the world's pre-minent military. CDAO is preserving that advantage by harnessing the best technology available and delivering it to our war fighters. Thank you.

[00:15:43]>> Thank you, Chairman Bacon, Ranking Member Kana, and members of the subcommittee. I appreciate the opportunity to speak with you about how DIU is accelerating the conversion of commercially derived technology into combat power. For 30 years, the Pentagon has favored expensive quality over quantity. The battlefield evidence is clear. While we are winning every engagement, we are losing on the cost exchange too often.

[00:16:10]Our adversaries have scaled production and reduced costs, forcing us to diversify our portfolio. We require both pricey high technology and lowcost lethal mass. President Trump, Secretary Hegsth, and Congress have directed an urgent rearmament, and Secretary Hegsth has tasked the Defense Innovation Unit with harnessing commercial companies to deliver accelerated combat power. Our determination is to dramatically lower our cost per kill while reducing risk to force. The good news is that America's unrivaled entrepreneurs and private capital markets have foreseen this rebalance. In recent years, over $200 billion of private capital has flowed into non-traditional and commercial technology companies. As the principal liaison to these commercial and non-traditional companies, DIU is bridging the gap from rapid prototyping to scaled successful transitions that solve operational problems identified by the joint force. We will be the fastest contracting and fielding agent in the department. We pioneered the competitive commercial solutions opening and other transaction authority to bring high-tech from solicitation to initial fielding in months, not years. As part of an innovation ecosystem coordinated by the chief technology officer to my right, we are tightening our linkages with the military services to deliver lethality they can scale. We're further knocking down transition barriers. These include classified infrastructure, authority to operate, testing and evaluation that delay technology delivery. Our basic investment philosophy is to focus at the intersection of speed, scale, and lethality to ensure the best return for the department and the American taxpayer.

[00:17:57]We thank you for your leadership from enacting long needed acquisition reforms in the fiscal year 2026 NDAA to the reconciliation funding in the one big beautiful bill. With your support, DIU will leverage the full power of America's markets and entrepreneurs to deliver combat power. Thank you.

[00:18:19]>> Thanks to all three of you. You have a very important mission area. We appreciate your leadership. Listen to Mr. Stanley's statement triggered a question to me.

[00:18:28]I'd like to ask all three of you if you could look at what you're doing at your accomplishments. Please tell us, but also the all the folks listening, what is an area or two that you've had a direct impact on the combat capabilities of our war fighters. And Mr. S, you touched on this a little bit, so we'll start off with you to give you guys a chance to think about it.

[00:18:47]>> Thank you, Mr. Chairman. Uh the the best example of that in CDIO is the deployment of Maven Smart System at scale across the entire uh uh global enterprise. Uh so at every single combatant command and most services as well as down to some tactical formations, we're actually using an AI enabled command and control capability uh called Maven Smart System. It's a software as a service uh deliverable. Uh but this uh this this capability allows for the integration of multiple different AI tools to accelerate workflows for operational decision- making to greatly reduce the amount of time it takes for operators to make better decisions faster on the battlefield. Uh this this tool has been widely used by just about every single uh formation that's been engaged in uh operations across the globe uh and is continuing to develop and build and get better over time. It's the first time in my career where an actual system of record has been used not only in current operations, but it's also evolving in real time with the expertise of our war fighters being inserted into that capability.

[00:19:50]>> Thank you, Mr. Michael.

[00:19:53]>> Thank you, Mr. Chairman. I'd say that the two things I'm most proud of um were really for the first time we created a pathway for new companies to deliver munitions. Uh and we did that by having our cross functional team evaluate the management team, the technology and their manufacturing capabilities. And we did that with all the military services and the other components. And we were able to pull them through the process in months, not years. And that means we get combat power in uh one year, not five years. The second thing I'm I think I'm most proud of is the innovation ecosystem, which was fractured. putting it together allows us to reduce duplication because this man on my right and this other man on my right, we could look at each other every day and make sure that we're not spending the taxpayers money uh in double or triple format. Uh and that that's also allowed the office of strategic capital and some of the other parts of the innovation ecosystem to do things fast like critical minerals deals uh rebuilding the defense industrial base. It's because we have a full view as to all the technology investments that we can make in the department.

[00:20:55]>> Thank you, Mr. West.

[00:20:57]Chairman, I'd first like to acknowledge my predecessors. Um, for decades, the department treated, you know, two exquisite pockets of American exceptionalism, the American war fighter and the American entrepreneur as distinct and separate. So, DIU exists to bridge that gap. And by and large, when I took the seat uh seven weeks ago, my predecessors had done that. Now, in terms of what to be proud of, the unit has many of these young companies. Emil mentioned a couple of them. I won't mention them by name. They touch DIIOU first in many cases. So some of the things you're seeing fielded uh these new entrance first touch DIU. Finally, to be specific and what's what Secretary Hegsth has ordered, it is a pivot to unmanned systems and DIU uh along with TRMC and other elements of the office of the secretary of war are running a program called drone dominance with a new template. It's an advanced purchase commitment uh which brings commerciality to the department um that that they that the secretary outlined during his acquisition strategy. Thank you.

[00:22:04]>> Thank you. It's ultimately about combat capability and what you're delivering.

[00:22:07]So I appreciate that feedback. Mr. Michael, I understand there was a reorganization. You were talking about it that by direction of the Secretary of Defense or Secretary of War. Could you We understand there's a a significant cut to your programs in the FY27 budget.

[00:22:21]Can you talk about that and maybe impacts uh to your portfolio?

[00:22:26]>> Uh I believe we asked for significant increase uh s uh Mr. Chairman across almost all items particularly what you'll see uh in next year's budget as new big items is something we call AI arsenal which is building the infrastructure uh required to do many many many more tasks by AI. We've also asked for increases for the mission engineering integration activity which is changing the requirements process. So we're not goldplating requirements.

[00:22:54]We're giving uh vendors a common operational problem that we're trying to solve and the S&T lines have been largely flat and consistent with sort of the department's budgets.

[00:23:03]>> I think I was to be more specific the military service science and technology programs. Can you talk about the impacts there? I you know I just got uh directive authority in the 26 NDAA to begin looking at the military services budget. So I haven't been able to impact those for uh for the 27 submission. But we are uh right now going through uh I hate to call it an audit but a review of all the technology spending by the services to make sure that we're spending enough in the right places and not duplicating where we don't have to.

[00:23:35]>> I appreciate your feedback when you get a chance to analyze the cuts and what the impacts are. And with that, I >> thank you, Mr. Chairman. Mr. Michael, I'm assuming you must be aware of reporting last month at the Guardian highlighting certain aspects of your disclosure to the Office of Government Ethics. The article highlights that your initial disclosure in March of last year of a significant amount of stock in Elon Musk's XAI, the company behind Grock, your subsequent sales of those holdings earlier this year yielded you about what percent gain >> uh about 5x.

[00:24:15]>> So about 400% to 4,800%.

[00:24:19]>> 500% >> 500%. In between you were serving as the Senate confirmed under secretary for research and engineering and chief technology officer. Uh under your purview was a $200 million contract with XAI to embed their Grock models on the department's AI systems. U I just want you to uh be transparent. Uh did you have any decision-making role in that uh XAI uh contract? No, that that contract was signed by the previous CDAO and that was done before that office was affiliated with Arie.

[00:24:56]>> So you uh recused your matters from >> it. It wasn't I didn't have to recuse it. It was in a totally different office.

[00:25:03]>> Did you recuse yourself from any matters related to XAI? Once the office ca the CDAO office which does the contracts with AI companies came under my purview in the fall after those contracts were signed in the summer I recused.

[00:25:18]>> Is there any decision- making or uh input you've had on XAI while you've been in the role of under secretary?

[00:25:25]>> Not until I sold uh my holdings.

[00:25:28]>> So while you had your holdings you had absolutely no role on any decision-m with XAI.

[00:25:34]>> I did not. And uh do you believe then that you addressed any conflicts of interest and were avoided?

[00:25:42]>> I do.

[00:25:44]>> So uh just to get fully your testimony, you never had any conversation with XAI.

[00:25:52]You never had uh any uh input on any decision-m with XAI. Uh and you had zero interaction with them while you were under secretary before while you before you you sold your stop. Yeah, I I recused myself completely from all matters regarding XAI. While I held that stock.

[00:26:11]>> While you held that stock, >> yeah, while I held that stock, I recused myself from all matters regarding uh XAI.

[00:26:17]>> Why not just make the recusal public then?

[00:26:21]>> I'm not sure I understand, sir.

[00:26:22]>> Have you made it public that you were that you had recused yourself?

[00:26:25]>> Well, what happened? Well, because my initial ethics uh submission when I took the role didn't have the chief digital AI office as part of the office. It wasn't my original recusal, but I contacted uh the our government ethics office inside the department as soon as I got the office and recused myself.

[00:26:48]But the bottom line is when you had the stock before you sold it, >> you had zero contact on any policy matter with XAI.

[00:26:57]>> Correct.

[00:26:59]>> Okay. I'm satisfied.

[00:27:02]>> Appreciate the direct answers with that.

[00:27:05]Mr. Finstead, >> I'm going to take a different route here. So, uh, thank you, uh, Chairman Bacon and Ranking Member Connor for, uh, having, uh, this hearing today, and thank you for the witnesses for being here. Um, I want to talk a little bit about the DIU's on-ramp hubs, and we are honored in Minnesota to have, uh, one of the newest hubs, and I had the opportunity to attend the ribbon cutting for that back in December, and I've already seen some positive impacts and heard some great things about what that is doing to help the Minnesota Defense ecosystem.

[00:27:36]So, with that being said, uh, Mr. West.

[00:27:39]Uh I frequently hear from small businesses with innovative ideas that are uh eager to partner with the Department of War across Minnesota. Uh however, I often hear of uh concerns about the so-called Valley of Death. I think Mr. Michaels, you addressed that a little bit earlier. Um sounds scary. I don't know exactly what it is, but I think we all should just instinctively avoid the valley of death whenever possible. Uh, so when these small businesses are struggling to bring their concepts out of the R&D phase and into the hands of our war fighters, that's that's what I'm hearing. So, how is DIU planning to utilize the on-ramp hubs to bring to bridge this valley of death uh by ensuring that these capabilities can be deployed across the department rather than stalling under initial demonstration phases?

[00:28:26]>> Thank you, Congressman. One of our statutory obligations is connectivity.

[00:28:31]And you know, we we frankly have to do a better job downstream of connecting with the military services to ensure that our prototypes and the young companies that that you mentioned have a home and some budget behind them. And so in many ways, we're a service organization responding to that demand signal. But we have to be upstream as well. And so well beyond Silicon Valley, the on-ramp hubs, you know, we have to look across the nation at all the great entrepreneurs, many of them veterans um in these small communities that do have the ideas, the engineering skill, the entrepreneurship to be harnessed and converted into combat power. Your specific hub, I mean, I think was opened in December. We've already seen, I think, 16 additional responses to our commercial solution openings, which are competitions that we run periodically. We've run about 20 so far this year. And, you know, candidly to to leave this off, DIU has the obligation to run these competitions as objectively as possible with with clarity and then sometimes to get to know early if if if something doesn't fit a military service. But we're certainly doing a disservice to these companies uh if we don't communicate well and and and create what what you described as the value of death.

[00:29:53]>> Thank you for that. Uh and I would invite you to Minnesota to to come see the hub and and to see these uh young young startup companies in some case seasoned mom and pop shops. You know, I represent a district that has an amazing amount of welders, tinkerers, inventors, thinkers, uh, companies that are starting to, uh, really, uh, tap into this ecosystem, and I think it'd be great to to have you get your feet on the ground there, eyes on the ground.

[00:30:17]Uh, Mr. Stanley, question for you. We uh in this committee uh subcommittee had a hearing not so uh long ago about you know really the uh the inventory and logistics and uh some of the you know archaic legacy software systems that are being used. And it just struck me when I hear uh the department of war using Excel spreadsheets and other uh you know tools to to track inventory. How are we really uh you know really putting our war fighter war fighter in a you know in a in a top priority space when when we have commercial technology uh you know big retailers they know where every toothbrush in their system is at every given moment but we're asking our men and women to go to an Excel spreadsheet to figure out where inventory is. So can you give me a little idea of how are we coming on that front and and what is your office doing to help you know really uh look at the commercial side and how do we bring that in?

[00:31:14]>> Yes, Congressman. So we're we're actually working quite closely with many aspects of this problem underlying all of the technology questions. It's really a fundamental data problem. How is the data stored? How is it collected? How is it managed? How is it curated? And how is it structured? so that you can bring some of those best of breed commercial tools into the department. Uh as uh the co actually reports to me as the CDIO of the department, our chief data officer, uh it's it's our it's our duty, it's our responsibility to actually enforce these standards. We have several data decrees that we have in issued within the department to get our data story right.

[00:31:51]Once you have the data story right, then you can start bringing in the tools. Uh we are also working very closely with members of the defense industrial base to try and improve the data structure and the data story visualization especially for exquisite munitions uh to try and accelerate not only their development and their deployment uh increase the capacity of of our ability to actually produce those munitions but also be able to track those more effectively especially for a protracted conflict.

[00:32:17]>> Thank you. Thank you Mr. Chair. Y >> you're recognized.

[00:32:21]>> Thank you Mr. Chair. Thank you all for being here. Um, Director West, appreciate your comments specifically about the legacy of DIU and the leaders before you. I think it's one of the things that is one of the greatest success stories, bipartisan, everybody working together, and I appreciate you saying that and the rapid progress you've made already in your time in the office. Um, no knock on my friend from Minnesota. Um but uh in my great state of New York uh we also have a great uh I think underappreciated actually defense technology um thriving ecosystem uh I know many of you have spent time there um recently and in the past um between capital startups um a lot of veterans and and military folks the fact that we don't yet have um at least a permanent defense tech presence or DIU presence in New York I think is opportunity and appreciate um the the conversations we've had and the work already in that direction. I also want to highlight um my alma mater of West Point in our great state that has made a huge investment at the academy um in building out an innovation ecosystem that draws on both the cadetses there but also the exceptional faculty, the labs and research space um that they have. and they're actually standing up an offampus commercial public private innovation works uh going forward that brings together partnerships with detachment 2011 Jiata 401 army futures the PAE's 200 plus industry partnerships gift space um so I just wanted to um ask your your thoughts on that I know you're you're drinking from the fire hose as you come on Director West but any thoughts there and any way we can partner to support that I think would be for the greater good of the nation Thanks, Congressman. I mean, I was I I worked for 20 years in New York City and and was part of the founding group of veterans on Wall Street. Um there's always been a an entrepreneurial spirit.

[00:34:20]I'd say the thing that has changed is capital allocation. uh as you pointed out you know defense tech is white hot now but we don't have to go back you know only a decade ago you had kind of people protesting affiliation with the department of defense and capital allocators uh were uninterested so New York um you know we have the requirement to establish criteria to judge where the next on-ramp hub will occur. New York, I think, objectively, uh, meets many of these criteria. You mentioned West Point. You know, I don't know what configuration this may take once we went run it through our evaluation system. I do know that West Point is also beyond the West Point works you mentioned. Uh, I believe they're starting a drone a drone range and you know, so that the Hudson speaks for itself as does the city. Um, I appreciate that and uh if you want to say on the record that it'll be the next hub, we'd we'd be excited to hear that here today or you know uh all joking aside, I know it has >> a little hard of hearing after >> um again appreciate the partnership um and uh go Army beat Navy on the record.

[00:35:34]Uh shifting to another topic, I um com commend you under secretary Michael your comments specifically seeing the launch of the low lowcost munition initiatives yesterday I believe or today whenever that was um exactly the right direction and it builds on um the drone dominance program that I know um several of you have worked on and just getting that high low mix correct. anything that we can be continuing to do on that front as a subcommittee and the whole of the committee. I know that's a a shared bipartisan um priority. Um I would love to talk more in depth about that, but I want to ask a specific sort of subcomponent to that building on what we've seen in Ukraine and and other uh um combat zones across the world. specifically looking at comm's networks and the opportunity for commercial cellular networks being used be for beyond line of sight C2 around drone swarms. Um I know DIU is moving towards Gauntlet 2 and three.

[00:36:35]Director Stanley um you guys are looking at Swarm Forge. Are you looking at integrating um commercial and private 5G options within that given kind of the complexities and and challenges in the modern battlefield? Uh, yes, Congressman, thank you for the question.

[00:36:53]Um, I'm I'm very much an egalitarian when it comes to comms, yes, as much as we possibly can. Uh, commercial comms in in 5G, line of sight, uh, laser comms, you know, smoke signals, I don't care.

[00:37:05]We're going to need the bandwidth on the battlefield regardless. And, uh, if there's a technology out there that enhances our ability to actually operate, especially in an EM contested environment, we're going to leverage that.

[00:37:17]>> Anything you want to add with the I guess we're out of time. All right.

[00:37:20]Thank you, gentlemen. Yield back, Mr. Chair.

[00:37:22]>> Thank you. Based on who is here, we >> Thank you, Mr. Chairman. Uh, Director West, in the FY25 NDAA, Congress established uh the shared commercial classified infrastructure pilot program, and we're now seeing amazing and extraordinary demand for shared commercial skiffs across the country, including in Colorado, where qualified companies and uh and research institutions are ready to support national security um but they lack access to accredited facilities and classified networks.

[00:37:59]Under your leadership, I think this program appears to be gaining real momentum with potential sites being actively evaluated uh prioritized and resourced uh including in my district of Colorado Springs. Can you walk us through how the department is identifying and selecting these locations? what steps you're taking to establish uh centralized DoD sponsorship, accelerate classified network provisioning uh and cut through the traditional accreditation and reciprocity reciprocity bottlenecks to deliver these capabilities at at speed and at scale.

[00:38:36]Yes, Congressman. I mean, we we've identified that um producing a prototype is insufficient uh when to hand that over to a military service which has its own bureaucracy.

[00:38:49]And so, as you point out, what you're going to see with DIU, and again, this this this started before me, um but it is being accelerated and and drastically accelerated. You're going to see a bundled service that includes skiff access, authority to operate, essentially can you bring software onto DoD networks and probably some access to boutique ranges or other places where we can test. In other words, we're trying to speed up the whole system. In terms of the evaluation of where these skiffs or secure compartmentalized facilities should be, the first thing we ask is, is this in a skiff desert? In other words, is is there a nearby or approximate skiff? Then we look at capitalization, the number of companies, the number of dual use companies, essentially usage or demand. Um, and and we we see a lot of demand. And so the list keeps growing, but I I think you're going to see this happen fairly quickly. As you said, there's wide department interest for many things like this under Secretary Hegsth, who has directed us to just cut through a lot of the bureaucracy, organize quickly. So this includes DIA, DCSA, ANS. Um there is large cooperation and and I think the underpinning is speed.

[00:40:08]>> Yeah, this is such a barrier particularly to smaller companies uh you know to to even entry in into the market. So anything we could do and I appreciate your leadership on that. Uh, Secretary Michael, last year DARPA terminated the Draco program, our primary effort to field nuclear thermal propulsion. Um, NTP would be a game-changing technology um to give the United States a leg up in the rapidly changing um battlefield of onorbit warfare. At the same time, the Space Force and Space Command continue to tell Congress that they need more tools to support dynamic space operations, including in space mobility.

[00:40:48]And General Saltzman's demand signals are pretty clear. The joint force needs the ability to maneuver satellites without the massive fuel uh penalties inherent to a chemical propulsion. At the same time, the CCP is aggressively pursuing its plan to field an NTP spacecraft in the next decade. Uh we can't effectively contest the CIS lunar domain if our assets remain sitting ducks in fixed orbits. Understanding all of this, how does OSW R& plan to continue supporting research into NTP and other technologies that support uh high delta V maneuverability?

[00:41:29]It's a good technical question uh for me uh but uh the DARPA program Draco uh uh served its purpose for in terms of delivering learnings. It's a very hard technical problem. The president's issued an executive order uh compelling the uh the whole government to figure out nuclear propulsion in space and the lead for that is NASA. So we're transferring the learnings from DARPA to NASA and then going to be working with NASA, the department of war uh for nuclear propulsion in space given the importance in the cy lunar uh atmosphere.

[00:42:05]>> Okay. I I have several other questions, but I I would go over my time if I did that. So I'm going to go ahead and yield back, Mr. Chairman. Thank you.

[00:42:13]>> Thank you there, Mr. Crank. Based on the order of arrival, Mr. Vinman, you're now recognized.

[00:42:19]>> Thank you, Chairman. So, uh, since we're all pitching on-ramp hubs, let me just pitch Northern Virginia, which is got hubs of military installations, including Quantico, Delgrren, um, big R&D bases, uh, great institutional, uh, educational institutions. So, uh, certainly give Northern Virginia a look.

[00:42:38]Um, now, during my 25-y year career in, uh, the Army, I've had sort of a throughine. I was a a land warrior platoon leader the first in the 82nd back in the late 90s and then worked on national security council as legal adviser to emerging technologies and then one of my last jobs was uh at DEVCOM army research and development uh basically 75% army's R&D so I think that cutting edge um technology and and um systems for service members uh the war fighters are absolutely critical So, let me start by thanking you for your your service to the nation, Director West, uh your leadership in uniform and particularly your tours in Iraq. I'm glad to hear that we're aligned on on the goals to ensure our armed forces have drones that they need to win the next fight or better yet, deter it. Uh last month, General Petraeus published an article in foreign affairs titled the autonomous battlefield. Not sure if you had an opportunity to review it, but uh ask unanimous consent to enter that article into the record. Without objection.

[00:43:45]>> Thank you. Um so I'll read a a quote. Um no joint doctrine um for autonomous formations yet exists. No major command has been tked to develop one. Uh no new unmanned systems force has been established. In essence, the US military is buying more drones without adequately considering how uh coordinated autonomous forces should be structured, coordinated, commanded, and controlled.

[00:44:09]that include uh unmanned systems in general. General Petraeus goes on to advocate for urgent development of joint doctrine to ensure we are prepared for to both train and use the systems. Do you agree with General Petraeus on the need for joint doctrine?

[00:44:24]>> Congressman, I do and I believe this is underway.

[00:44:29]>> Can you uh comment a little bit more about that, Congressman? I I did read General Petraeus's article and and many of his basic points uh the the Department of War uh is enacting and really have been for several months now. But in terms of the new budget requests where 54 odd billion dollars has been requested for autonomous systems, baked into those lines are uh doctrinal and schoolhouse monies to do exactly what you're you're indicating. Okay. I look forward to seeing the doctrine on that because uh obviously I think uh we uh in the United States autonomous systems uh unmanned systems are very nent and um uh some of a lot of that development has been going on in in Ukraine for several years and they're much further advanced. So um modern battlefields especially in conflicts such as Ukraine have demonstrated the vulnerability of traditional radio frequency control drones to jamming and electronic warfare. From DIU's perspective, how significant is this challenge for US forces?

[00:45:36]>> Congressman, when we uh launched drone dominance, the first thing we did was ask for military experts. You know, as suits, we're trying to tamp down the bureaucracy and move as fast as we can with a commercial buying program. And what those military experts uh said is is is I think where you're going. And I know you're very deep on this subject, but the the the three basic conclusions that they've drawn from the Ukrainians specifically, it's first the fusion of the commercial world with the battlefield and the frontline units.

[00:46:08]Second is the the drastic reduction of cost per kill. Uh and third is is speed and the fact that this is an iterative when I talk about speed, these iterations determine who the victor is monthto month and sometimes week to week. And so what you're implying is there's an offensive and a defensive component that are intertwined and EW is a inherent component on both sides. So one solution is the use of fiber optic guided drones which is uh you're well aware is resistant to electronic warfare and capable of operating GPS and RF denied environments. How do you assess the operational and tactical value of such systems for US forces?

[00:46:49]I think a a diversity of systems is required but it but as you say uh well let me give a a very specific example.

[00:46:56]We recently held our first gauntlet which is a competition um over 300 sorties. Five Ukrainian uh drone companies participated in the Victor used fiber optics. American companies are now on the move and and we know of several that are going to integrate this in the next gauntlet.

[00:47:15]>> Great. Well, I hope that DIU can play a a significant role in in prototype and validating and accelerating these uh technologies. And thank you. With that, I yield back.

[00:47:24]>> Mr. McGuire, you're recognized for five minutes.

[00:47:26]>> Thank you, Mr. Chairman. Thank you to our witnesses for being here today. I've got to respond to my colleague and say, "Go Navy, beat Army." But I love them all. Of course, what you guys do is very important. Uh there are many ways to serve. I served in the SEAL teams and it's amazing how much technology that the new SEALs are getting every day as we constantly update. Software seems to have been something that was probably at the bottom of the top 20 priorities and now software is closer to the top because things change as you can see in Ukraine on a day-to-day sometimes hourly basis trying to keep up and Mr. West I agree with your comment. The fiber optic uh is a great idea but also you need to look at all the options. Um so Secretary Michael uh your office has consolidated the list of critical technologies from 14 to six. How will this streamline development and capabilities in these new technology areas and how will this affect your work on technology areas that aren't on the list anymore?

[00:48:24]>> Uh yes, Congressman, we basically the idea or the thesis around um how we thought about critical technology areas which are a mandate from Congress. you ask us to tell you what we think the critical technology areas are are those things that are not being focused enough on by the department where there is focus even though it's still critical we just it's not on the list for example space is of critical domain for us but we have a space force a space command it's a wellunded set of uh in our budget request so it remains critical but not on the list the things that we aren't investing enough in are the things I added to the list Uh so it doesn't mean that those things that fell off are not important. It just means that we have uh momentum in the department for those items.

[00:49:12]>> Thank you. Uh Director West, in last year's work and families tax cut act, we provided over a billion dollars to supercharge the small UAS industrial base. How often do you get to talk to the actual operators on the front lines who will use these technologies?

[00:49:28]Congressman, we've we've integrated the some of the best operators in the military services into the competition.

[00:49:35]So, the operators are designing these competitions and then they judge them uh by flying them off often with just two hours of preparation for a new system to determine if that system can be scaled for for average drivers like myself.

[00:49:51]>> And I know we we're working hard on the drone dominance program. Working with these operators on the ground is probably helping you get fast feedback to make better, quicker decisions. Uh, Mr. Stanley and Mr. Michael, yes or no?

[00:50:03]Is there one large language model that does everything every user needs?

[00:50:09]>> No.

[00:50:10]>> Okay. Uh, yes or no. Is it important to be able to use large language models across all classification levels?

[00:50:18]>> Emphatically, yes.

[00:50:19]>> Yes. And uh what is the department working on to allow access to multiple large language models across classification levels from one interface?

[00:50:29]>> I'll go ahead and and take that. Um so when when it comes to large language models, the integration of those models into a number of different domains uh all classification domains that works ongoing. Uh we have jai.mill that began in the uh cui uh classification. um that really was our first foray into the deployment of a large language model for the entire enterprise. We are trying to replicate that success on uh our secret and topsecret domains. Um that work requires more bespoke engineering as you would expect uh for for the deployment of those very specific models and capabilities at those different classification levels. That work is ongoing. We are leveraging all of our lessons learned from our uh you know class or my uh CUI deployment uh for both secret and top secret. Uh we're looking for similar capabilities that we have in in CUI on on secret and top secret. But most importantly, we're trying to leverage a variety wide variety of models uh across all all of those domains so that we can have the best technology for answering the best questions for the right purposes in real time. I was in the Pentagon last week and I saw posters with the Secretary of War encouraging Department of War to use AI. Is do you is there do you think there's a hesitancy for folks in your career field to use AI?

[00:51:46]>> In my career field uh no sir there's no hesitancy. In fact, in my office, I I I stated if we're going to profess that we are an AI first uh department, which we are, uh everything in CDAO is going to be identified and use AI at speed and at scale, we are going to do or practice what we preach every single day.

[00:52:04]>> How can we in Congress help you to cut the bureaucracy in the red tape so that we can speed things up because things move so fast with software upgrades and everything else? What can we do further?

[00:52:18]Congressman, I think that's probably better taken uh in a in in a different setting. I don't think I can answer that question in five seconds.

[00:52:25]>> We really appreciate your service, guys.

[00:52:26]Thanks so much. I yield back.

[00:52:30]>> Thank you. We're all catching on here.

[00:52:32]Miss Hulahan.

[00:52:34]>> Thank you, Mr. Chair. Uh thank you for the chance to say uh to have questions for you all. I'll start by my plug, which is for education. Full stop. Um, I would love it if I could ask of you all some of the ways that we can accelerate and talk about only AI and figure out what technologies a seal needs is to educate people and to give them a fine quality education. And it seems as though you all I believe come from Harvard and from Stanford and from themies and the fact that we are in a environment right now where those are being maligned and uh abused I think is atrocious. So I would ask of you all that you defend the power of our education system. Um I come from Philly.

[00:53:14]We have a lot of universities and a lot of research there. Um and that will be my ask of you uh for for now. Also u my questions have to do with biotechnology.

[00:53:25]I spent a lot of time in that space. I worked a lot of time on the NDAA the last NDAA on putting in some of the commission's recommendations uh therein and Mr. Mark, I'll some my questions will be to you about that and I'm not sure to the degree that you've been able to come up to speed on that. But in the NDAA, we established the biotechnology management office um which is hopefully going to organize and advance and elevate biotech across the uh agency across your agency. Was wondering if you might be able to provide an update on that.

[00:53:56]>> Yeah, we're very excited that uh that you put that in legislation. We've appointed a leader, a gentleman named Gary Vora. Uh we've actually written the charter and we've submitted to the congressional staff to get uh feedback from from Congress as to is it hitting the right notes relative to the legislation and it's uh sitting at this the deputy secretary's office so it's got the right uh level of importance and elevation in the department and it's synchronizing with my uh biotechnology critical technology area. So it's really going to bolster our efforts and thank you for doing that. That's awesome and and really what I'm really glad to hear that and look forward to possible opportunity to meet with him. Um how do you think the department then will be able to ensure as the law did did intend that the biotechnology management office will be fully funded and staffed and empowered? Is there anything else that you think that you might need to make sure that we are keeping pace at a minimum and uh maybe even better than that with China? I I think uh what I'd like to do, what they generally like to do is give the leaders a little bit of a 30 days to get in the seat and then um uh take the charter feedback from Congress, combine that with the leadership mandate and then come back to you with some uh some questions or suggestions if we need to bolster it in any specific way.

[00:55:14]>> Terrific. Thank you. uh also related to the NDAA and biotech section 247 asked the secretary of defense or war to create guidelines on the ethical and responsible development and deployment of biotechnology within DoD or DOW to better support critical innovations while letting the United States do what it does best which is leading by example through uh strengthened norms. Do you know what the status of these guidelines are or that who and who is leading that effort? Is it the same Gary uh Batvana?

[00:55:43]Okay, Gary Vora. Um, I'm not sure. I'll have to get back to you on that, but I think um, you know, one of the reasons I made it bio, we call it biio manufacturing, uh, part of the critical technology areas and why I think it's going to synchronize well with the office that you created is that we're going to be able to to really move faster relative to China. They've invested a lot in this space and we have to do it in a way that's consistent with our values and principles. And I think uh that's going to have to come out of this joint effort um at the department.

[00:56:13]We'll get back to you on that.

[00:56:15]>> Terrific. And kind of related to uh the conversation on biotech and biommanufacturing and also uh my plea to you about academia. Um how are you guys at this point engaging with academia and with industry and with local communities frankly as well to make sure that we're incorporating their perspectives as well into the uh NDAA uh guidelines that are required.

[00:56:38]We I mean we know R& has probably the most extensive university partnerships of any component of department of war. I meet with the university presidents uh very you know almost uh every couple of weeks. We have strong partnerships with John Hopkins with Stanford with a a lot of DoD research DO research dollars go into these programs. Um, and also something that maybe uh, director West can address is we we're really in need of cyber talent coming out of these universities. So, we're very focused on growing a next generation uh, of talent that wants to be technologists both for the government and outside the government. Uh, uh, we're part of we're a big part of the tech force. uh if you've heard of that which is recruiting university students to work in all different parts of the government uh to bring their tech skills to us and then they go back to the private sector. So uh >> I'd love to have a longer conversation.

[00:57:32]I've run out of time on that and on um cyber education and cybermies and cyber everything. Um and I yield back. I appreciate that.

[00:57:42]>> Just had to reshuffle the order here real fast. Mr. McCormack, you are now recognized.

[00:57:47]>> Thank you, Mr. Chair. Great to have you all here on such a a great occasion and especially with such an important topic.

[00:57:55]I think I spend about 80% of my time on this topic. I think between cyber and the future of what's going to make us hum versus what's going to get in the way of progress. Uh this is one of the more important uh one of the more important topics especially from where I come from.

[00:58:10]Georgia Tech is one of the leading universities in America and technologies. Um Mr. Mitchell or sorry, Mr. Michael, sorry. Uh, I'm I'm excited about this year's budget request, but I hope that we keep in mind that the award winning technology of the future comes from basic research done at places like Georgia Tech, which I just mentioned, uh, which has enjoyed decadesl long research relationship with the department across all levels. How will the department ensure the basic research programs remain a priority so that we aren't caught with inferior technology in the future conflict?

[00:58:46]You know, you will see from uh from the office of secretary of war side consistent promotion and uh of the continued growth of basic research dollars going to universities. We're expanding our partnerships very consistently uh across the landscape, not just at the Ivy Leagues, but at state universities and technical universities. Um and I think with the new authority that uh Congress was was kind enough to grant me to direct some of the other military departments to ensure that their budgets are appropriately fun financed at that level as well that we'll be able to do that in the coming years even better.

[00:59:23]>> I like that premise. However, I was curious what is the reasoning behind decisions like eliminating the Navy's university uh research initiative?

[00:59:32]>> I'm not familiar with that sort of it was a Navy program.

[00:59:35]>> Yes, I I'll get back to you on that. Uh obviously we have University of North Georgia that's won a competition on on cyber security. That's another good university in my area. I tell you I'll switch over to Mr. Mr. West. Uh how is the realignment of your organization under research and engineering impacted your authorities people mission and ability to collaborate across the service and defense innovation enterprise?

[00:59:58]>> Thanks Congressman. I I think the the summary word is alignment. Um Emil as the CTO is the architect who sets the innovation strategy including for the military services now as a field activity. Um you know I I feel that this DIU has more durable authorities and we still have operational independence. So Emil sets the framework. We have strategy meetings as do the military services once a quarter. Um, and with within this framework, we hadn't lost any speed.

[01:00:32]>> Okay, Mr. Stanley, I'm going to pose the same question to you as far as the the organization of the research and engineering uh impacted your authorities, people, mission, and ability to collaborate across service and defense, innovation, enterprise. In other words, is it getting the way or is it enhancing?

[01:00:49]>> It's definitely enhancing, representative. Uh so when when I look at uh where we were before CDAO was aligned underneath uh research and engineering that was before my time as the the CDIO but I I have experience inside of this architecture this ecosystem uh from from my previous roles in the department. Um what what I can say emphatically is I've ne I've never seen CDA more empowered or enabled or connected to sister organizations across the entire enterprise. Uh I am one phone call away from director West or from uh director uh Wshell over in DARPA. Uh there are many examples I can talk to where AI is such a foundational capability and foundational technology area. We need to coordinate much better.

[01:01:33]Being underneath a single uh chief technology officer has greatly enhanced that.

[01:01:37]>> Okay, this one will be thrown to the general public. What can we do better?

[01:01:41]Where can we uh streamline the process?

[01:01:43]Usually it's the government getting in the way of progress. How can we do better?

[01:01:49]>> Well, I uh we have to continue to make it easier for newer companies to succeed while having existing companies be held to account. Um, and you see that a lot the deputy secretary's actions working with the defense contractors to change our business model, their business model and ours. So, we are not goldplating requirements, but we're holding them accountable on time and cost and really got to welcome the new entrance in and a lot of that's our own bureaucracy getting them through the process so they don't get stuck at qualification and testing, getting them time zone ranges, uh, giving them clear demand signals, quick yeses, quick nos is what I call it. And a lot of that's on us, but some of that we could certainly use your help.

[01:02:31]>> Yeah, that's always been the crux of the defense problem is that contracting issue. I appreciate you guys being here and being part of the solution rather than the problem.

[01:02:39]>> Okay, we had votes called, but we got time to get our remaining two folks. So, we'll do that. Well, Mr. Whitide, you're recognized.

[01:02:46]>> Thank you, Mr. Chairman. Um, good to see you all. Um, thanks for uh we had a nice meeting yesterday and uh appreciate the time. Uh I I wanted to just start by saying that I I believe that the areas in your purview are the heart of DoD innovation and thus the heart of our national security. And I appreciate uh all of your stepping back into government service um to do what is a crucial crucial thing which is to pull this massive enter enterprise that you're in charge of now into the speed and scale that is required uh to meet the challenges of the moment. So first I wanted to say thank you. Um Mr. West, my first question is to you. So, I appreciated you coming to my office uh a few weeks ago to discuss how DIU is accelerating deployment of innovative technologies. Can you describe how your organization was able to use advanced purchasing commitments to fasttrack the drone dominance project? And of course, uh Emil, if you'd like to add, I'd love your thoughts as well.

[01:03:47]>> Uh certainly, Congressman, a year ago, it sounds astonishing, but we didn't have a a weaponized drone program in in the department. So, as Secretary Hegsth gave the order and Congress was good enough to devote monies to autonomous systems, when we designed the program, the advanced purchase commitment stood out as as we examined past mistakes on having a steady demand signal so that young companies and and of the to give you an example of the 25 finalists who competed uh 23 of them were were new entrance or or startups. s um but was lacked in the past is this steady demand signal. And so with a billion $1.1 billion dollars sitting out there by using a competitive process, these companies could build product road maps.

[01:04:40]And then as we've gone through these gauntlets, the other thing that's been illuminated are supply chain constraints. And then our our brethren in office of strategic capital and other areas of the department can now lend money against our supply chain. So that I am convinced that that a year from today we'll have a healthy American supply chain and we'll have drones that can compete on any battlefield in the world and that and that starts in our judgment with advanced purchase. So there's no ambiguity about the money that's on the backside.

[01:05:12]>> Yeah. I I will add that you know one of the advantages that a company applying to DOU has is that they're theoretically supposed to have a dual use. So they have another avenue for revenue. There's a whole another category of companies that are only building for the defense enterprise and that there's a huge venture capital push right now toward those companies. If we can't get them through the department, which means giving them an advanced purchase commitment, which says if they build something that works, that that we can't commit to them over multiple years so that they can continue getting investment dollars, it's going to be tough to sustain uh the defense tech industry that's very excited right now about the new pathways we're creating for them.

[01:05:53]>> I think I thank you. I I think I speak for the entire commit committee or at least I hope I do and when I say that we want to be good partners to you um sir as we uh navigate this big challenge ahead and um so if there are additional authorities or other ways that we can support uh the crucial work of innovation um we want to do that and uh and I know the chairman and I have talked a lot about that. Um, the last thing I'll say is just my pet issue that I always raise in every area. As you know, uh, California and the American West is um, uh, in a wildfire crisis and that threatens our military installations. It threatens, um, our readiness. Vandenberg Space Space Force Base almost burned down uh, a few years ago. Uh I would appreciate your partnership as we work to make sure that our bases are uh protected and uh and uh that we protect our our troops and our people and I think in doing so we'll also make big investments in the American uh west both in terms of autonomy and in sensing. I don't know if there's anything that you want to add to that subject. Yeah, I believe that uh DARPA is collaborating with the state of Texas on uh on how we could use autonomous aircraft to to to treat wildfires. Um and I believe DIU is looking at some commercial capabilities in this area too. It's a you know it's not a state problem. It's a national problem. We think about that way because uh because the impacts have have ripple threat ripple effects through the economies. Um so we take it seriously at the DARPA side. I don't know if Owen wants to address what the you said.

[01:07:28]>> In the past, we we've worked with the California Army National Guard. Um, you know, you're on point for this issue, Congressman, and I think it's a combination of commercial dual use, as Emil has said, and then AI uh in terms of recognizing and responding. Um, but but you you do have our commitment to to cooperate with you.

[01:07:47]It's this is a pretty obvious one. Thank you for your service. I yield back.

[01:07:52]>> Miss Jacobs, you're recognized.

[01:07:53]>> Thank you, Mr. chairman for letting me wave on to your subcommittee today. Um, thanks to our witnesses for being here.

[01:07:59]Um, I want to talk about DoD directive 3000.09.

[01:08:03]Um, it remains the department's core policy on autonomy and weapon systems and requires that autonomous and semiautonomous weapon systems be designed to allow appropriate levels of human judgment in the use of force and that they undergo rigorous testing, validation, and evaluation.

[01:08:19]But the directives additional senior review requirements for autonomous weapon systems without human supervision does not cover operational decision-making beyond the scope of an individual weapon system. So platforms like Maven Smart system that you Mr. Stanley uh mentioned earlier they sit earlier in the chain of decision-m helping to identify classify and surface objects or potential targets at scale and that can materially shape the lethal outcomes even though a human being makes the final decision. So, um, Mr. Stanley, what department policy ensures human responsibility over the decisions to use force since directive 3000.09 only covers specific weapon systems?

[01:09:00]>> There's a variety of policies, uh, Madame Congresswoman, that actually dictate how lethal force is going to be used throughout the department. These policies very much predate any AI integration into the department. Um and they are standard practice for all of our war fighters and the operational commanders who ultimately have the responsibility to leverage military capability in any any operation. Uh they those are the ones that dictate what decisions are made, why they are made and how they are made. They are very robust uh especially when it comes to again any type of operational decision- making with the use of le lethal force.

[01:09:37]>> Okay. Um, so what so why isn't there like a 3,000.09 or equivalent for the use of artificial intelligence in operational decisions?

[01:09:50]>> The operational uh decision-m process is is is quite robust when it comes to you not only the people but also how the data arrives, what data is actually leveraged in order to make those decisions. And it's also quite varied depending on the operation, depending on the geography, depending on what we're trying to do in in as far as accomplishing our objectives. Uh so so when you look at uh the the overarching policies, again, I I I would yield to my uh OSW policy colleagues as far as the questions about individual policies, but when it comes to the integration of AI, we view that as any other technology.

[01:10:25]How are we improving? How are we becoming more effective? how we're becoming more efficient in the application and delivery of those capabilities in real time. Uh that's one of the reasons why we are focused so much on how we are deploying AI especially directly to those warf fighters. They're the experts in how to use any type of military decision- making. And by partnering with them, we get the perfect combination of human decision- making with topshelf military or uh AI enabled technology in order to integrate them into a a complex, you know, uh application of that human machine team.

[01:11:00]>> Yeah. Yeah. I just want to make sure we have robust protections in place for for human judgment. Um, Under Secretary Michael, is the department considering any changes to directive 3000.09, especially any changes that would weaken oversight over decisions surrounding the development and fielding of autonomous weapons without human supervision?

[01:11:19]>> Uh, I'm not I think it's a policy directive, so it's not in our our group, but I'm I'm I would expect that all kinds of policies are changing as we learn from what's happening in Ukraine and Russia. Um, and we learn about the capabilities of AI and we learn about what it can and can't do. I think those policies are ever evolving. So, I wouldn't want to commit that they're never going to change, but I'm not involved in any change uh right now.

[01:11:41]>> Okay. Thank you. And for either of you, as DoD integrates AI tools into the targeting process, how is the department incorporating civilian harm mitigation into the design, testing, validation, and fielding of those systems?

[01:11:56]When it comes to civilian harm, again, that's not the responsibility of R& uh we we are working very closely with our counterparts there to make sure that they understand the implications from a technical perspective. Uh as well as integrating all of these technologies into the right robust framework for the operational commanders to know what their limits are. Uh but when it comes to the civilian harm mitigation that that again I I defer to my colleagues in in in policy and PNR. So as you all are evaluating these tools and deciding what to sort of bring in, are you at all looking at the civilian harm risks every day? Congresswoman, my the objective of our technical integration approach is not to uh limit what we do on the battlefield. It's to improve the decision-making process. And what we found is that AI by itself makes mistakes. Humans by themselves make mistakes. What we endeavor to do is take the human cognition, the human discreet uh the the ability, our decision-making processes and augment them as best possible with artificial intelligence, not only to make decisions faster, but also better decisions faster to minimize as much as possible collateral damage on the battlefield.

[01:13:11]>> All right. Thank you. Thank you, Mr. Chairman.

[01:13:13]>> Thank you. I appreciate the testimony of all three of you. I think you can tell you have bipartisan support here. That that's my sense. Uh folks have trust in you. You're credible. You're experienced. Also appreciate the direct concise answers. Uh we don't always see that I may point out. So we're grateful to you. And with that, we'll close the committee. We'll go vote.

[01:13:34]>> Thank you.