Science and Technology Priorities in FY27 Defense Authorization Request and FYDP

Good afternoon and thank you for appearing today. The subcommittee welcomes today, Under Secretary of Defense for Research in Engineering and Chief Technology Officer of the DOD, Emil Michael, Deputy Assistant Secretary of the Army for Research in Technology, Mister Chris Manning, Deputy Assistant Secretary of the Navy for Research, Development Test and Evaluation, Mister Dave Tremper, and Associate Deputy Assistant Secretary of the Air Force for Technology, Engineering and Product Support, Doctor Janet C. Wolfson. Thank you all very much. Our witnesses appear before us at a critical moment. China, Russia, North Korea and the weakened Iranian regime continue to threaten US interests, our forward forces and the homeland. This subcommittee has heard extensive testimony on the challenges we face, from China rapidly closing the gap in artificial intelligence to the risks posed by adversarial biotechnology. Today we are here to examine how the Department of the Service and the services are working together to research, develop and field the capabilities our warfighters need faster than our adversaries. In Ukraine, we have seen the timeline from innovation to shrink from months to minutes. US small businesses can play the same role for our military, and I look forward to hearing how the department is leveraging their innovation and accelerating the transition of new technologies to the force. I also want to understand how lessons from recent conflicts are helping reduce the time it takes to deliver capability to the war fighter and how we can help move even faster. Our service members performed heroically during operations against the Iranian regime, and I would like to hear how the department's science and technology enterprise contributed to that success. From the first combat use of the Lucas drone to advances in AI-enabled systems, the United States cannot afford to fall behind in harnessing the innovation and entrepreneurial star strength we have here at home. On the topic of AI, there are two areas I would like understand. First, I want to know where we stand when it comes to harnessing the best of what America's frontier labs have to offer for our national defense. These labs offer our war fighters a key advantage on the battle space, and I want to understand how we are engaging with them to make sure we are integrating them to the greatest extent possible. Then, second, I want to discuss leveraging autonomous systems. Thank you, and I will now turn to the ranking member. Senator Slotkin.

Thank you, Senator Ernst. Um, and sorry to make you wait there at the beginning. We were both voting so that we didn't have to interrupt you all again. Um, I'm um super interested in this hearing and really appreciate everyone being willing to come and talk about this I know that it's um often complicated to do that in an unclassified setting. Um, we had a classified briefing on the classified AI race, particularly vis-a-vis China uh not so long ago. But I think it's important that we start to have this conversation out in public, um, as much as we can, because the public should be invested in understand that this is actually a race. Um, that, um, uh, uh, similar to races we've had in the past, such as the n- the nuclear race, um, who gets to the most capable AI first does really matter, and I think people need to understand that, um, as they struggle to understand what AI is gonna be just h- Um, as I understand it, the Chinese are about six to twelve months behind us, um, which um, sounds like both a little and a lot, depending on how well you know this topic. Um, and I just, I think it's hard to overstate how AI is gonna change how we fight wars, how we think and engage in conflict, who wins and what the definition of winning is. Um, and unlike um warfare of the past, um, this um new kind of wave is being led by private sector companies. It's not um government funded labs or research institutions that are coming up with all of this stuff. It's the private sector, um a number of companies who are cutting edge, and you all in your various departments um have to manage those relationships and make sure we win even though you don't control the things that go on at this private I really believe in our potential to win the fight. Um, and I I just will not give in to people who have a like a dark cloud over, um, this race, but I do think it's gonna come down to people, chips, and power. And, um, making sure we have those key ingredients, that we have the best people and ideas, the most capable chips, and we have power, meaning energy, in order to, um, support that future. Um, I will um save my questions uh till our multiple rounds here, but I appreciate everyone and what you do. Um, we are all invested in a bipartisan way in being successful, so thanks for being here.

Yes, thank you, ranking member, and we will start now with opening statements from our witnesses. We'll start with uh Secretary Michael and we'll just proceed across the the dais here. So if you would, you're recognized for five minutes, Secretary.

Thank you, Chair Ernst and Ranking Member Slotkin and uh distinguished members of the committee, when they get here. Um, I'm honored to appear along uh my esteemed colleagues, Doctor Janet Wolfson, uh, Mister David Tremper and Mister Chris Manning. Uh, my mandate mandate as the department's chief technology officer is to ensure delivery of tangible decisive and battlefield-ready technology to our military. and to do it at velocity and scale. Uh, the urgency stems from the hard truth that if we do not maintain a dominant technological advantage American service members will pay the price in combat. The reality is hardened, is heightened by the strategic environment we face. Uh, our adversaries are investing heavily and moving with the purpose to surpass us in key technology areas that will define the uh the course of future conflicts, including AI, quantum, munitions productions. Um, however, we have a s- uh, strategic national asset that our adversaries don't. We have a dynamic innovation ecosystem, uh, comprised of unmatched talent and unlimited creativity and lots of capital markets that support those industries. My office remains focused on harnessing the full potential of our innovation ecosystem to serve the American warfighter. If you remember, after I was confirmed last May, my first order of business was restore focus on technologies that we need to maintain dominance on the battlefield. I inherited a list of fourteen critical technology areas, which really didn't prioritize anything. They were a laundry list that diluted our resources and inevitably slowed progress. Through rigorous data-driven analysis and in consultation with the military services and experts in the IC, I made the decision to pair that that list down to six key areas. Those include applied artificial intelligence, biof- biomanufacturing, contested logistics technologies, quantum and battlefield information dominance, scaled directed energy, which means cheap and proliferate, uh, proliferated, and scaled hypersonics, which also means cheap and proliferate, proliferated. The department is focusing in its, its investments, its talent, its leadership around these areas, while the private sector crowds in capital to supercharge the process. I wanna be clear that the other technologies such as space and microelectronics remain important R and E will continue to carry out our statutory mandate to provide rigorous oversight uh across the department's R and D portfolios, uh and sure attention, advocacy, and resources continue to be devoted to those areas. Um as an example of this quick progress, we rolled out Gen AI dot mil, which was the department's first uh department-wide ai platform which didn't exist before and uh in just a few months we've got one point four million of our three million members use have used uh our gen ai tools which is an incredible increase from the eighty thousand before that but we're not limiting that deployment on class networks alone which was what gen ai uh started we signed agreements with virtually all leading ai frontier models infrastructure companies in the last few months to deploy their capabilities directly onto the department's classified networks. Um, we are also bringing innovation to production and acquisition of conventional weapons enabling the return of mass effects through the employment of large numbers of less exquisite weapon systems. From the deployment of Lucas drones, which were able to try in theater and speed up the R and D process to make them a program of record We've also signed deals uh with new entrants in the last week with five new companies to develop low-cost containerized cruise missiles and low-cost hypersonic missiles. Uh, these are uh an evidence of our openness to new industry that will do fixed price, fixed cost, and fast delivery. So as I noted earlier in our uh my remarks, uh the innovation ecosystem now has a unity of effort in the uh Department of War, we've consolidated the Chief Digital and AI Office, DIU, DARPA, the Office of Strategic Capital, SCO, and TRMC under my leadership, and that gives us the ability to reduce duplication and increase effort on the things we need to as a department. With that, I'll leave it to my colleagues.

Thank you very much. And, Mister Manning, you're recognized for five minutes.

Thank you. Chairman Ernst, Ranking Member Slotkin, and distinguished members of the subcommittee, thank you for the opportunity to testify today on the Army's innovation enterprise. It is an honor to represent the dedicated scientists, engineers, and acquisition professionals who work tirelessly to ensure that our soldiers remain the most lethal, protected, and technologically advanced land force in the world. This is an exciting time for the Army as we embark on a generational transformation of how we do business. The battlefield of today is hyper-connected, highly contested, and characterized by extremely rapid technology cycles. To dominate in this environment, the Army cannot rely solely on acquisition methods of the past. We must fundamentally transform how we discover, develop, and deliver capabilities to the tactical edge. In coordination with the Department of War and our partners in Congress, the Army has undertaken a bold acquisition reform effort, aligning the systems of how we innovate, develop, and buy under portfolio acquisition executives. This reform ensures that every dollar spent on our applied science and technology enterprise is deliberately synchronized with requirements generation development and fielding. It allows us to get new, over-matching capabilities out of research and into the hands of our soldiers more efficiently and effectively than ever before. Last year, in coordination with Army leaders, we conducted a comprehensive review of the Army science and technology program, identifying areas where the Army must keep or increase investments a- and areas where we should rely more on commercial sector innovation. As a result of this analysis, we began to realign funding through reprogramming, and you can see the full effect of this rebalancing in our FY twenty-seven budget request. We increased S and T investments in areas such as long range fires, counter C five ISRT, and counter UAS. We plan to capitalize on heavy commercial industry investment in areas such as C three and autonomy. Overall, the Army's S and T budget request for FY twenty-seven is two point seven seven three billion uh about eleven percent above our FY twenty-six request. The Army faces competing needs which we work thoughtfully to balance. We must provide our soldiers today with advanced capabilities to meet the current threats without abandoning the long-term fundamental research necessary to maintain our advantage for the next century. Our FY twenty-seven budget request acknowledges this reality. You will see we made reductions to our basic research program. We increased our funding requests for applied research S and T. We remain committed to a robust level of fundamental science. We have also continued to improve and expand the Army Research Risk Assessment program. This effort safeguards taxpayer-funded research by ensuring Army investments flow to academic partners, who are both scientifically credible and main- capable of maintaining secure research environments. As we've seen over the course of multiple conflicts across the world, product development cycles can be measured in weeks and days, not years. To address the need for extreme agility, the Army shifted how we engage nontraditional defense contractors, start-ups, and commercial tech sector. We are taking the best practices of technology maturation in commercial industry and applying them to how we do business in the Army. The newly launched Pathway for Innovation in Technology, or PIT, represents innovation in the dirt and operates under three guiding principles. Speed is more important than perfection, early soldier iteration is critical, and we must validate the demand signal before we make large-scale investments. The pit ties together army service component commands, army corps, and army divisions with the industry engagement efforts of army fuse. With this construct, the pit allows acquisition professionals to iterate directly with soldiers in the field, turning tactical demand signals into fielded prototypes in months rather than years or decades. While this new program is still working towards full operational capability, it is already starting to bear fruit. Finally, on behalf of the entire Army S&T community, I would like to express my appreciation for the critical support this committee provides. We cannot succeed without you. Thank you again for the opportunity to testify today. I look forward to your questions.

Thank you, Mr. Manning. Mr. Tremper, you're recognized for five minutes.

Chairwoman Ernst, Ranking Member Slotkin. Uh, thank you for the opportunity to appear here for you today to address the state of the Department of Navy's science and technology budget portfolio. Department of the Navy S&T budget is essential for maintaining future naval power and national security amidst rapidly evolving global threats. The budget is aggressively focused on transitioning critical basic research to deliverable combat capabilities while ensuring a high return on investment through disciplined, relevant, and rapid technology development and integration. The strategic landscape is transf- transforming with relentless technological change and a new era of great power competition. Basic and applied research are instrumental in creating the scientific foundation for emerging naval concepts, technologies, and capabilities. Advanced technology development investments in future naval capabilities evolve that scientific research into new transition-ready combat capabilities aligned with naval operational priorities, while investments in innovative naval prototypes translate that research into revolutionary naval combat capabilities. The Don has recently established the Department of Navy Rapid Capabilities Office as our single accountable organization for accelerating the fielding of critical solutions to the fleet within a three-year time frame. The Don RCO is intended to fail fast and iterate quickly with industry to ensure capabilities are tested and refined in operationally and relevant environments. The Don RCO brings operational challenges directly to industry. This enables a parallel focus on new entrants and traditional primes to the defense industrial base, but also harnesses American innovation more effectively to deliver the best possible solutions to our war fighters. The Don is also actively supporting defense innovation ecosystem transformation efforts spearheaded by Honorable Michael in order to strengthen alignment across science and technology research and development organizations. Relatedly, the Department of Navy is working to improve internal alignment of naval research and development establishment to best enable the new program acquisition executive's centric acquisition and uh enterprise strategy. Optimal alignment will ensure timely delivery of capability and operational support to so- to the Fleet and Marine Corps while ensuring long-term health and stewardship of the NRDE. The NRDE depends on world-class facilities to develop and deliver capabilities. This infrastructure has been challenged by facilities that are exceedingly surpassing their traditional life expectancies with limited new investment due to other higher priority needs across the department. I am thankful to Congress for recognizing this challenge and providing authorities that bring new and enhanced tools we can leverage. I am furthermore supportive of the lines of effort established in the FY twenty-seven annual five-year plan on improvement of Department of Defense innovative inf- infrastructure that highlight additional opportunities to strengthen RDTNE labs and facilities. An effective NRDE enables successful transition of emerging science and from the laboratory into the hands of our sailors and marines. Key to that success is our ability to quickly translate scientific discoveries into and across naval operations systems and platforms. Department of Navy is proactively aligning our activities across acquisition reform, requirements reform, and innovation ecosystem reform to streamline capability delivery and empower PAEs with flexibility to integrate new technical solutions, as those solutions emerge. Critical to this accelerated capability delivery is seasoned technical workforce within our warfare centers which can shepherd S and T development within or outside the the department of war to deliver survivable and sustainable capabilities within naval operational environments. Don is working to optimize access to and operation of our warfare center's engineering activities as a critical enabler for transition of innovative S and T and industry solutions into the hands of our war fighters. These efforts reflect the acquisition transformation strategy, rebuilding the arsenal of freedom as directed by executive orders and authorities granted in the fiscal year twenty twenty six National Defense Authorization Act. But Don is proactively aligning our activities to realize this acquisition requirements and innovation ecosystem reform, while promoting competition, incentivizing faster execution, taking calculated and shared risks, and leveraging new authorities to create acquisition flexibility to deliver the decisive

Thank you very much. And now, uh, Doctor Wolfson, you are recognized for five minutes.

Thank you, Chair Ernst, Ranking Member Slotkin, and distinguished members of the subcommittee. And thank you, uh, for the opportunity to testify today on the Air Force's fiscal year, twenty twenty seven, science and technology budget request. We are proud to represent the Air Force's science and technology enterprise, working in full alignment with Secretary Honorable Michael, uh, to ensure the United States military and especially the United States Air Force maintain a decisive and overwhelming advantage in any and every domain of warfare. The Air Force Research Laboratory's mantra of " one lab, two services" captures the integrated nature of our mission, supporting the critical air, space, and cyberspace domains. Our fiscal year twenty twenty-seven budget request is directly responsive to the department's priorities making increased investment in critical technology areas, and de-conflicting our efforts across the services to focus on Air Force unique technology gaps. We are committed to delivering tangible, battlefield-ready results for our warfighter at velocity and scale. To achieve this goal, we continue to invest and prioritize in investing in decisive military capabilities. Our fiscal year twenty twenty-seven budget is focused on delivering the capabilities needed to deter conflict, and decisively prevail in any battlefield scenario. The following investments represent some of our commitment to fielding those technologies that provide an advantage to the American war fighter. First, as a top modernization priority for the department, the Air Force is rapidly advancing offensive hypersonic strike capabilities and the layered defenses required to counter adversary systems. We are also focused on the materials and manufacturing advances, needed to field those systems at scale, ensuring our warfighters have the tools to project power and hold any target at risk. Another area is ensuring dominance in the air domain, and is increasingly increasingly reliant on resilient and high-performance energy solutions, such as advanced battery and energy storage. While the commercial sector continues to drive standard battery innovation, the Air Force is deliberately investing where industry will not, targeting the hard United States Air Force battery challenges inherent to extreme combat environments. We are prioritizing advanced specialized battery technologies to enhance the performance, endurance, and reliability of critical operational systems from high altitude unmanned systems and platforms to forward deployed exp- expeditionary equipment. Investing in nuclear delivery science and technology is also paramount, as all Department of War operational plans depend on a credible and reliable and effective nuclear deterrent. The United States Air Force's nuclear delivery science and technology programs, focus on developing and re-establishing the specialized expertise, infrastructure, and nuclear d- delivery system capabilities needed to modernize and mature the science and technologies that will enable the Air Force to operate in a peer-contested nuclear environment. Finally, the ability to fly, fix, fight, our most advanced platforms is paramount to maintaining air superiority. The Air Force is addressing the sustainment challenges for our fifth generation fighter fleet across the entire platform, from canopies and airframes to advanced engine components. Crucially, we are optimizing fleet readiness by advancing non-destructive investigation capabilities. By investing in these technologies that improve reliability and reduce maintenance burdens we ensure our most critical assets are ready to project power, and defend our national interests. We are also revitalizing the innovation ecosystem, accelerating the links between our organic science and technology enterprise, the commercial industrial base, and our acquisition programs. These enhancements to our technology transition and innovation ecosystems are essential to our efforts to evolve our acquisition engine which will enhance our ability to identify, evaluate, and recommend promising technologies for early adoption. Every investment we make in every technology we accelerate

Thank you again to our witnesses and we will start now with uh the rounds of questions and answers and i'll recognize myself for five minutes. And uh Secretary Michael we'll start with you. Uh you had stated that uh bio-manufacturing is one of the six critical technology areas that you had personally selected. In last year's NDA Congress mandated uh the creation of the Biotechnology Management Office to ensure that the department can catch up to China. And that was an important step, but um a mandate if we don't have that leadership prioritization, it really is just an organizational chart entry. So we really need to see this move forward. Um, Iowa State University, of course, um, my alma mater, is home to one of the department's bio-made pilot facilities, and it's scheduled to open in twenty, twenty-seven. aligning with the department's broader sprint timeline, and we really feel this is progress worth protecting. Uh, but GAO has found that the department plans to wind down two of its three major bio-manufacturing initiatives by fiscal year twenty-seven and fiscal year twenty-eight and the congressionally mandated biotechnology road map isn't expected until later this year. So, Under Secretary, how will you ensure the Biotechnology Management Office becomes a genuine priority? And how does the department plan to sustain momentum in bio-manufacturing beyond uh those program end dates?

Uh, thank you, Senator. Um, we have recently announced a leader for the Biotechnology Office reporting directly to the Deputy Secretary. We've drafted the charter, we've shared it with uh, I believe your staff. Uh, so that we can get comments from congress and do we think that we have the right legislative intent embedded into the charter um we also have the critical technology area in my group that will support that biotechnology office so we think we're putting more resources into to this and not making it just an organizational checkbox

mmm thank you

um in terms of checking on the bio uh the two facilities i'll have to get back to you on that i don't have the answer off the top of my head why those decisions were made

ok

Um, but we take this very seriously, I believe it's a threat vis-a-vis uh the progress that China's making on some of these technologies. Uh, and that's why it's one of my critical technology areas.

No, and I thank you for that, because when we talk about um e- emerging threats and capabilities, uh we don't often talk in the bio space. It is really important that we do so, um because if we're not engaging, you know, China is just going to get that much further ahead. So I appreciate that and look forward to continuing to working with you uh on that area. Um we've also seen for more than a decade with uh the department of war that there seems to be an incessant ability of senior leaders to really ignore the importance of infrastructure. And it's especially true when it comes to our labs. Uh yet the department's only idea for for fixing this is to raise the unspecified minor military construction threshold. And this is troubling considering the needs within the department are hundreds of millions of dollars to uh modernize that infrastructure. So for all of our witnesses, uh why do you think the department is failing to program for lab infrastructure? Is it competing requirements? Maybe just please go into that and um Doctor Wolfson will start at your end of the table, and work our way across. And I have a minute and a half remaining, so if you could be brief.

Excellent. I think the challenge really is the prioritization of near-term needs versus long-term uh prospects. A lot of the lab infrastructure is not for this fight, but the next fight or the fight after that, which becomes challenging. The m- uh advancement to major MilKom is or minor MilKom is actually a huge step forward that allows us to use our flex four funding um for up to fifteen million dollars and mol- many lab buildings and technology can be uh uh invested with those little bits of sums. In fact, for AFRL we have um a hundred and thirty-six million dollars that we have allocated for a flex four for next year that we could utilize that additional ceiling for.

Okay, thank you, Mister Tremper.

Thank you.

I I will also double down on what what um Doctor Wolfson just just pointed out. The the flex four funding has been critical for the infra- for us investing in our infrastructure, and I will also uh completely agree that operational capabilities today and MilKom needs today, in addition to barracks and families and child care things,

Mm-hmm.

are outweighing um our long-term uh S and T investments in infrastructure, and so we find ourselves behind and but extremely appreciative of those authorities that we're getting through Flex four and even section two thirty-three that has allowed us to start investing in our infrastructure.

Thank you. Mister Manning?

For the, for the Army it's similar, it's prioritization uh regarding barracks, family housing, daycare, and then where do we get two labs? Again, I'll echo section four forty one twenty three that flex four has been extremely helpful. We appreciate Congress uh providing that ability for us to do improve our facilities.

Mm. OK. And any additional thoughts, secretary?

I'll just say that um in the recent NDAIA you gave uh Congress was kind enough to give uh our office the ability to do some oversight. So that allows us to work in tandem to look collectively of where we can get efficiencies uh with the spend that we need to make to upgrade the labs from the state that they're in, which is not great.

OK. Thank you. I appreciate the feedback. Senator Slotkin.

Um, thank you, thank you for all being here. Um, you know, I I think that if we were if our politics were healthy, this body would be dealing with AI and the coming impacts of AI every week. We'd be having some conversation about the economic impacts, the military impacts, the intelligence impacts, the way that it's going to change um our country, I think, merits real conversation and there's an opportunity cost for us being, you know, up here fighting about the daily issues. We're we're missing some of the big strategic moves. Um and um Secretary Michael, can you help us understand and kind of in English for the public, since we are in public, why is it important to win the AI, classified ai race against china what would happen if china beat us to the punch can you put that into words for us

yeah i'll start with the national security implications and maybe some of the economic implications but from a national security stand point um this is another case of uh our adversary the main adversary china you know taking our ip from our american developed labs that have spent hundreds of billions of dollars by the end of the next couple of years between the cloud layer and the ai company layer and they're they're quote unquote distilling those models which means effectively copying them for a fraction of the price taking off the guardrails for them which means they could be used in ways that they're not intended to be used which is very dangerous for us whether it's cyber as a cyber weapon as a biological weapon as a chemical weapon uh and when they can take those guardrails off and point essentially that ai against us uh that's a very dangerous imbalance that i worry about a lot um and when it comes to using ai to develop new weapon systems um they have a culture which permits or they can direct the government the chinese communist party can direct their industry their ai industry to build whatever they want uh we have to get on a better page with our domestic uh partners in the ai industry to make sure that we're developing unique capabilities when it comes to physics aerodynamics material science all the things that could go in to make our defensive systems more lethal more accurate uh uh and so on and then the last piece is from an intelligence stand point they're willing to use ai um from an intelligence stand point in a way that we haven't contemplated yet and that gives them greater insight we've seen intrusions from salt typhoon into our phone networks, into s- individual phones, into our systems, um, and that could be sped up by AI and made more intrusive. And so the threat is real, that we have to stay ahead, uh, of our - of our adversaries in every dimension of AI. And that's part of why I believe the AI action plan to - to make AI American companies stronger is really important.

Yeah, I mean, I think, um, You know, this is the problem when we're we're dealing with an adversary who doesn't have the rules that we so desperately believe in, um, in a democracy and w- are trying to preserve. Um, but, you know, how am I supposed to square, um, you know, we wanna beat the Chinese to the punch, but they don't have any guardrails and we need guardrails, right? I, that is, i- it is a, it's a conundrum. Um, But, you know, I I introduced a piece of legislation on guard AI guardrails for the military, not to not to, um, unduly stifle development, but just to say that we should have a human being in the conversation where we're making decisions of life and death. We shouldn't launch nuclear weapons without a human being, uh, making that decision, and we shouldn't, you know, have domestic st- surveillance. All things that the Chinese could decide to do tomorrow. So help me square how we keep our sort of values as a nation, but still win this race.

Yeah. Well, we ha- so we have to stay ahead on chips, power, innovation, and capital formation. And that keep, that gives us this six to twelve month lead, and maybe we could extend it. In the last commerce NIST evaluation, our lead had increased. by a few months against the Chinese so uh and part of the reason for that is to try to prevent this IP theft that's happening, that's a big deal, this distillation attack thing is a is a real problem and we've coordinated as an administration with all the labs to try to help them prevent these distillation attacks, we also have to prevent them from uh help them prevent uh operational security mishaps, there's a lot of foreign nationals that work at american ai companies that we have to be concerned about and have to help them secure their own networks on and we have to work with them also uh on ensuring that the the way that um uh there is a flow between researchers from our country to theirs um doesn't sort of uh you know release the secrets that we don't want uh done but it's going to be an ongoing challenge there's no one solution to this

yeah i'm i'll have more to say but i yield back and my time's up

Thank you. Senator Shaheen.

Thank you, um, Madam Chair, and thank you all for being here. Um, Under Secretary Michael, the cross-functional team for anomalous health incidents is now under your portfolio. And as we've discussed, I think, before in the full committee, um, many of us who have followed the AHI issue over the past, really, more than six years now, have had concerns that the shift um places a more of a technical focus on an issue that really deals with not just care coordination but intelligence attribution research heavy coordination with the inter-agency especially at a time when I understand there's breakthrough research um related to attribution and I was really surprised I have to admit to see that the budget request um nearly triples the amount that we've seen in past years for the cross-functional team. So can you give us a little more detail about what, where you hope the focus of this funding will be, and what you're hoping to accomplish?

Yeah. Um, I think the right place for this effort is, uh, in research and engineering because, uh, diagnosis and treatment for the victims is priority one. Priority two is attribution, and then three is policy. And we have to get the first part right to be able to do two and three. We're not gonna, we can walk and chew gum at the same time. So, our inter-agency collaboration has only increased since it's gotten to my team. We've gotten a large, uh, you know, the first increa-, ma- major increase in collaboration from the IC on this. Um, and we've also followed congressional guidance for the first time where we have a two-star Navy SEAL m d as the military deputy on hihi so we've staffed it up with some of the best and brightest minds the uh budget increases so that we can do more of the real world testing because we believe that um if we can confirm the science that's preliminarily been confirmed over the last few years we can prevent we could develop countermeasures we can do the treatments better uh our databases will be stronger and we'll know how to teach the doctors at Walter Reed how to diagnose and treat better. So I think all those things are all good news stories for where this is now, and that's why you see the budget increase request.

Well, I

Well, I look forward to the briefing, um, tomorrow and,

Yeah.

uh, to hearing more about what you're doing. That's encouraging and hopefully, um, we'll answer some of the questions that we've had for a long

I think you'll be pleased by the briefing.

Thank you. Um, I wonder, um, I wanna also go to the SBIR program, something that, um, is near and dear to us in New Hampshire because, um, Senator, former Senator Rudman, um, introduced and passed that legislation and it really came from some of the innovation that was being done by small businesses in our state back in the eighties. And it's really been, it's been a critical tool, but there have been some changes that have mandated that agencies implement a cap on the number of proposals a company can submit. I understand that Deputy Under Secretary Ma- Mazal, am I pronouncing the name correctly? Told the ranking member of the small business committee that the cap would be implemented on a per topic basis. Is that your understanding as well?

Yes, on a per topic basis so that um, you know, w- we thought the total cap on any entity was different than capping on a per topic basis, which had some benefits in reducing some of the abuse of the system. but gave everyone an opportunity to participate.

Um, great, I probably could have asked the chairwoman that question as well,

Yes.

since she chairs the small business committee. Um, but I'm glad to hear that that's the way it's being interpreted. Um, my final question is really both for you, Under Secretary Michael, and for Doctor Wolfson, and it has to do with PFAS destruction technology, um, because as you know, these have been really, um, numerous at U S military compounds, um Portsmouth, um the former Pease Air Force Base was one of the first bases in the country to show, um PFAS in significant amounts in the ground water in ways that affected um children in particular. And there was a provision in last year's NDAA to authorize the use of new destruction technologies. So can you give us an update on um what the department is thinking in terms of PfAS destruction and disposal and maybe Doctor Wolfson I'll ask you to go first

uh that is an excellent question and not in my area of expertise so I'll take that as a QFR and get back to you with that answer

ok then I will ask Under Secretary Michael if he can respond to that

I'm in

I am sadly in the same position I'll have to owe you an answer on that man

ok well Good. I look forward to having those answers to the subcommittee. Thank you both.

Yeah, thank you. And I I do wanna uh recognize Secretary Michael as well, um, providing feedback for the SBIR STTR reauthorization and we uh really saw the department respond very quickly after the reauthorization and there were over ninety topics that were brought forward uh by the departments for competition, so Um, thank you very much for that. Now I recognize Ranking Member Reid.

Oh, thank you very much, uh, Madam Chairman, and I've got three basic questions I'd like to address to all of you. Uh, one, you're talking about your priorities, but I don't think you've put in your FY twenty-seven budget request for the record, and that would be useful. So if you could tell us what your budget requests are and a comparison with last year, are you or you're down. And also, uh, some of the areas where there are major funding increases and decreases. So, Secretary Michael, and run down the list, please.

Uh, it's been a confusing year for budget requests given how the budgets have gone with, uh, so I know you resonate with that,

Yes, sir.

uh, Ranking Member, but, um, we submitted ours to - to our - our leadership at the Department of War, so we've done all the work on, at least at the component level. Uh, I don't know if it's with OMB now, but we've done, uh, at a high level. uh on the rdt and e side defense wide from my office we asked for uh about a billion dollar increase from last year um at the core at the core rdt and e level um and that's to fund certain programs in hypersonics and some of the key areas in the critical technology areas mmm separate from that you'll see two big requests on um compute power which we need for the ai arsenal essentially that we need to build And for uh, an autonomous warfare group, where we're going to uh, really coordinate our autonomous, which includes drones, counter drones, and all the things that we need um, to fight sort of the next war, not the last war.

Right. And those requests are gonna be armed.

Those are gonna be what, sir?

They're gonna be recognized by OMB and included in

Those, we, uh, yes. We, yes.

Good. Mister Manning, please.

Uh, thank you, Senator. Uh, The R I'll start with FY twenty-six. So in FY twenty-six the Army requested two point four eight billion dollars for Army S and T. Of our FY uh our FY twenty-seven request w- uh we submitted is two point seven seven three billion dollars, so about eleven percent increase. I do want to note that uh Congress enacted four over four billion dollars in twenty-six for S and T so we thank Congress for the bipartisan support to Army S and T.

Right. But you your confidence you're gonna get that eleven percent increase?

Uh, yes, that's what that's what we've submitted in the in the budget.

And then uh where's the big plus ofs and the big decreases?

Yeah, the the big increases were in long-range uh fires uh counter UAS and counter C uh counter C five RSRT. Uh, areas where we took risk in our leverage planning or intend to leverage commercial investments uh are C three and autonomy.

Mr. uh Tremper please.

Uh, yes, for the Navy. So in our our requested was uh two point five five uh billion. In FY twenty seven our request was two point four zero eight billion, so approximately five percent decrease. Yeah. Uh, part of that included a realignment of six three money into six four. So when I talk about S and T, I'm talking about six ones through six three. There was some realignment of six w- six three investment, which was more appropriately aligned to six four, so that was removed. Uh, the, there's a significant emphasis right now in that integration space. Right, and so that word integration becomes extraordinarily important for taking S and T and putting into the hands of the war fighter. So you start to see budget adjustments that are happening between the six one, the six three, and then emphasis happening into the six four, which is where we do our prototyping and we get it into the hands of the operators. Um, so, uh, the enacted for FY twenty-six was three point two nine nine billion, so significantly more than requested and greatly appreciated those ads that that we benefit from that. Um, our fight up profile for our BA one through threes has been pretty consistent within that, that ten percent timeframe.

Uh-huh.

Um, but really we're seeing a significant increase in the integration space on how do you take it from the lab to the warfighter.

Right, so that's lab, warfighter, and command and communication, C two through three.

Yep,

Okay.

exactly.

Doctor, Wolfson, please.

Yes, so the, for the United States Air Force, and again I'm only speaking for the United States Air Force, if you want the answer for the Space Force we can take it as a, as a QFR and bring it back for you. Um, our FY uh twenty-six uh presidential budget was two point five three four uh billion dollars um and for uh FY twenty-seven uh our request is two point seven eight billion dollars so we have gone up we're focusing uh significant impr uh increases in hypersonics um we left our directed energy line um where it should have been, where it has been and we took some near term risk in bio-manufacturing and autonomous AI for uh collaborative platforms. Uh, we're also, though, very grateful for the eight hundred and thirty, uh, million dollars in, uh, congressional interest items to support the S&T community. And thank you very much for that.

Quite right. And, uh, I fondly rec- recall the Air Force as the US Army Air Corps, but forgive me. Thank you.

Thank you for joining us, ranking member. Um, we'll go ahead and do another round, um, and I just wanna double back on the labs for just a moment if each one of you could share maybe what is your number one unfunded priority when it comes to our lab infrastructure secretary michael if you have input please first

i mean most of the lab infrastructure spend is gonna come from the services so i think it'll be worthwhile to start with them yeah

just go ahead mister manning for the army thank you

uh thank you senator uh each year we submit our section twenty eight a twenty eight O six report of our unfunded lab infrastructure.

Mm-hmm.

Uh, I would go to that and go to the top of our list, uh, to be able to provide it. So, I don't have that with me specifically here.

OK.

I'll be happy to take a question for the record on that.

OK. Thank you. Mister Tremper?

Uh, yeah, I, uh, so for our unfunded priority list, uh, our number one is a reconfigurable cyber security lab at Nioic Pacific, uh, to address cyber security concerns concerns. However, equally important to that is our number two, which is an electromagnetic and cyber countermeasure lab at the Naval Research Lab to support our counter C five ISRT investments as spectrum electronic warfare,

Mm-hmm.

um sensing our key priorities within our S and T strategy going forward.

OK, wonderful. Doctor Wolfson, thank you.

Uh, for the Air Force we were lucky enough to make it onto our unfunded priority list. We are number eleven, um and it's a human performance uh S and T lab on the DAF priority list. However, I will also reinforce what Mister Manning said. Uh, all the rest of our infrastructure, uh, that did not make this list is listed in that same report, and that we provide to Congress every year.

Okay.

Thank you.

Thank you very much. Um, we'll go back to A- AI and autonomous systems. Um, I think this is something that's really important to both, uh, the ranking member and I, but, uh, we do know that the department has been making major strides when it, it comes to AI. and autonomous warfare, and I think large in part not just this subcommittee but the entire committee, uh, really supports the direction that we're heading. But as we're seeing the program scale then, the policy, just as, uh, Senator Slotkin has has pointed out, the policy architecture really has to scale with it. And this is where we probably lag behind. But the department is requesting a massive increase for dog, um for those in the audience that may not know, is the defense autonomous warfare group. And it's going from the two hundred and twenty five million up to the fifty five uh billion dollar mark for fiscal year twenty seven. And at the same time we're integrating the AI driven targeting with those autonomous munitions at a pace that DOD directive uh three thousand point O nine uh was not designed to contemplate. Uh, so, um, secretary Michael, has the department formally reviewed whether the current governance framework is actually keeping pace with dogs' growth? And then how do we overcome that if not?

Uh, I th- I think um, I think we have to evolve that. Um, uh, and I think the policy part of DOW takes the lead in uh amending or updating three thousand nine, and we'll be part of that. but it absolutely needs updating not only because of the capability potential increase, because of the threat environment, what's possible by the adversary, uh, and w- uh, but partly because of the lessons we learned in Iran. You know, we want autonomous mine, uh, uh, seeking, um, capabilities. We want m- autonomous mine, uh, defeat alternatives. Golden dome has an autonomous element to it, a space-based interceptor. that could take a high uh you know to hopefully get a chinese hypersonic missile in the first ninety seconds of launch before it separates into decoys and multiple munitions so there are going to be different risk levels with autonomous and we have to account for them in our policies

mmm

and my my belief is that will change more frequently than it has in the past and it ought to has to be consistent with our values consistent with the threat environment and consistent with the technology development

yeah Thank you, I I absolutely agree with that. Um, we understand the urgency. We want to provide proper oversight of that. We don't want it to be taken as um opposition to uh obtaining our goals. Uh, but we also understand that China and Russia, they have few less governance issues, uh, when it comes to autonomous systems and AI than we do, but we still do want to align. to our values. So, um, I really appreciate that discussion and I will yield my time and will go next to the ranking member.

Thank you. Um, yeah, ju- I mean, I think what you're hearing is, is that there's a lot, there's a lot of questions about how this governance should be done. Um, and I am just keenly aware that unlike the nuclear race, right, who can get to a nuclear weapon first? Right, the Oppenheimer story, um, where that research and all that funding and all that support and that development that went on in the middle of the desert to get us to a nuclear weapon was paid for by the US government sponsored by the US government not a private sector company. And fast forward to today, the private sector is really where the just powerful innovation is happening Um, and it, I think it puts even more onus on all of us to have that conversation separate from each company um on what the governance should be for AI and I, I do not believe that an a private sector company should get to decide what the rules are but I gotta be honest I think it is part of our congressional role up here to provide left and right limits um that provide some guidelines for how we govern this very new technology, and just like Congress did in the nuclear age, right? It wasn't like um we just said, " Hey, private sector, go bonkers. We now discovered nuclear energy and nuclear weapons. No oversight needed from Congress. Let us know how it works out." It was given that it had the potential for life and death, it had real oversight from this body, with input from everybody. But I just I I feel very strongly and I I think um also in that time, right? It's not like before we started the arms control conversation, I trusted that the Soviet Union would always do the right thing with nukes. Of course they were willing to do things that we weren't willing to do, but I'm not willing to sacrifice our values with this next generation of technology and just say well the Chinese are gonna do bad things towards us, so we should say no rules. And I was very glad to hear Um, actually that President Trump and Xi Jinping actually did seem to talk about some sort of guardrails on AI even generally at the summit. Um, but I just wanna go back, Under Secretary Michael, to your comments and the importance of winning the race, right, cuz I again I think there's also bipartisan support for winning the race, um, and how important it is to get there first. Um, and um, your concerns about the Chinese and distillation and kind of

Mm.

stealing our AI models or learning and and taking so it's you've d- you've laid out how important it is to win you've laid out how they're stealing um um given the strategic importance of winning I cannot for the life of me understand two decisions that have been wa- made number one the the decision excuse me to sell NVIDIA chips to the Chinese um giving them not our most sophisticated but some of our sophisticated chips, and chips they do not have. Um, and then secondly, I do not understand picking a fight with one of the few companies, Anthropic, that's in all of your systems, all of you use Anthropic right now, um, to the point where we've named them a supply chain risk and all of you are supposed to be divesting from Anthropic in the next two months. Help me understand, um, given what you said about IP theft, why we're selling them some of our most sophisticated chips, and why we're kicking out one of the top three in, you know, AI companies from all of your classified systems.

Yeah, so, um, on the chips question, um, the, uh, this is a bit, so this is a debate within the technology industry, which is, um, if you sell an adversary, uh, older chips, do you slow down their domestic production of equivalent chips because they become reliant on your technology. That's one theory. And then the programming language on those chips to the developers, which are really important in AI because they develop on top of these chips and on top of these technology stacks. If they become used to the American stack, is that net better for the American AI prolif- proliferation? And that's a, that's a debate and, and the White House has decided that if we gave them two versions behind chips, that we'd be able to preserve our our dominance on the programming language and um, make it less encouraging of them to develop their own domestic chip industry to catch up.

But do you have any evidence, based on what you said a minute ago about their theft, their rampant theft of IP,

Mm-hmm.

their rampant ability and desire to, you know, work backwards from our technology, from cars to drones to chips to whatever, What evidence do you have that giving them this technology ad gets them addicted, rather than um they just steal it and reverse engineer it for their own purposes?

Well, it's, you it's very hard to re do, reverse engineering in this particular category of chips is not uh possible given sort of the steps in the chain. You have TSMC, the chip manufacturer, you have the lithography machines that may there there's a a whole chain of things that they would have to do to be able to reverse engineer and replicate that. Um, that being said, I think, um, the number of chips also matters, if you think about what one data center in the US has, the biggest one has two hundred and fifty thousand chips in one data center. Uh, the kind of chip numbers that we were talking about in total for export for a count- country that's four times our size in terms of population was less than that, in one data center. So this wasn't it's not a sort of broad hey go sell everything to anyone at any time and numbers matter in this in this race as you said before the amount of electricity the number of chips how many data centers we have are they interconnected and does the data flow in the right way to them so i think this it could be that we're making too much maybe of this one um sort of fig leaf if it was um but i'll uh wanna if you're open i'll go to the second question

I'm I'm over time here, maybe we can come back to it and we'll and we'll come back on on traffic.

We'll come back, we'll come back.

OK.

Thanks. Yeah, thank you. Senator Peters, you're recognized.

I'm sure. Actually, I'll probably come back to that question uh so you'll have to free my time you'll be able to answer some of uh Senator Slutkin. But I wanna talk uh primarily right now about um our our need to deal with uh drones uh and swarms. We had a discussion about some of this uh in my office, but probably one of the uh key areas is uh deploying uh directed energy. weapons is probably the the way you're gonna be able to to deal with it so if you could just tell the committee a little bit how you plan to to mass produce uh directed energy in this uh field cause it's uh um you know you can have drone versus drone warfare other kinds of uh countermeasures but at least what i have been seeing and in test is directed energy is gonna be probably the major way in which we can uh deal with these weapons so how do we accelerate that i think we're way behind right now how do how do we catch up uh and make up for our lost ground

yep i think um a few things that we're doing and remember uh scale directed energy is one of my top critical technology areas so it is a focus uh uh by us we the science for directed energy is largely done and now we're in the engineering phase of it so to engineer it the engineering part of it makes it cheaper smaller and more proliferated Um, and we've now have a suite of directed energy, uh, products that go from low-end to high-end. And now we have to scale production of those. The things that are helping that are the, our Golden Dome, because they have a big reliance on directed energy. Um, certainly our experience in Iran has also, um, doubled our interest in these systems. Same thing with high-powered microwave, although it has a different sort of mechanism, fashion than lasers. Um. But we've also developed a joint inter-agency task force, four O one, to collate all these different technologies so that we have bigger demand signals, which encourages the industry to produce more and we get scale economies on that.

But, but let me just jump in uh,

Sure.

so you, I mean, the, the DOD has set a time-line of thirty-six months for, for uh accelerating this.

Mm-hmm.

That's like light years compared to what we've seen in the past. What's different? What, what, how are we gonna hit that? Hopefully we can do even shorter than that, but What are we actually doing different?

The Yeah, so there's a lot of the money allocated to Golden Dome is going to go to the fundamental engineering of these systems so that we can make them cheaper small and smaller and more proliferated. So that's a big change. Um, and because the commitment was made to the present that we're gonna have a demonstration that includes directed energy in our Golden Dome architecture, there's a lot of energy going into that.

When do you expect that to be?

while we're gonna have multiple demonstrations, the primary demonstration where it demonstrates a lot of capabilities will be summer of twenty eight. Um, so we have a whole team on this. There's never been more effort in the department on this particular um, capability. Uh, and, you know, uh, there are several companies that are emerging that have developed it and several companies that are taking what they've already built and making it cheaper and better.

OK, then I'm gonna follow up on uh, Senator Slotkin's. question, 'cause um uh we've talked about this in the office as well,

Sure.

is that uh we have to have access to all of these frontier models. We know what with uh Mythos now, uh Mythos uh coming out was a game-changer in so many ways, but that's not gonna be the last big one that comes out. There're gonna be others that are gonna come out. You know, how do you take certain companies off the table? We we need everybody on the table, everybody engaged. We shouldn't be calling some companies uh security uh risk and supply chain risk. when they're the actual leaders in the world, we should be figuring out ways to make sure that we're working closely with them. Uh, I I have lot of questions about what we've talked about in the office, lot of questions how that happened, but

Sure.

give us some assurance that uh we're gonna be making sure that our cyber warriors have the absolute best tools and take advantage of all the amazing American companies that are coming up with incredible new technologies.

Yes. Uh, well, seven out of eight companies we've signed up with, before I got in this seat and I I had the responsibility we had one partnership with one company now we've got seven partnerships including open source models which are another vector that are very important that are not close source um and all of them have agreed to the dow terms which are all lawful use cases, only one hasn't and just to be you know be specific like I was in your office for the other members here the things that we currently can't do under the contract with that that company that we declared a supply chain risk We are prohibited from doing battlefield management applications, prohibited from doing weapons development, prohibited from directing interceptions, like the golden dome uh scenario I gave, I gave you, or prohibited from targeting military infrastructure even of an adversary in a conflict. Uh, so those terms don't work for the warfighter. So maybe the technology is good, but if the terms of use don't match and they could shut the system off, if it's used in a way that they don't like, that becomes a non-viable product for the warfighter. That's why it went out and made sure that uh, OpenAI, SpaceX, Microsoft, AWS, NVIDIA, and Reflection AI are all partners on DOD on classified networks. We have the best of all worlds.

So I'd like to follow up because, uh, you mentioned areas there that was not been out there in public, uh, as to what those limitations are. You know, you've any

discussed that we discussed in your office and we're working through how we could get you there's some NDAs between the vendors on that contract that i'm reading from uh these terms so that if if if there's a way to do that we'd like to make it do you to see

Well, I definitely wanna see that. Is that something we can do?

Yeah, I I believe it's something we can do, but we have to just work with our OGC on that.

Uh-huh. When when will I be able to see it?

Yeah, I wou- I was trying to get it to you. I just didn't know we'd have to go through some legal processes, but I think we can do that as soon as possible.

I have I have I have a public commitment now, so that's great. I appreciate that.

it's a public amendment for me, I have out brought it by the lawyers but yes, but I I I want you to see it. You you know, uh Anthropic could give permission any time to let you see it.

Okay. Good, we'll talk. Thank you. Thank you, Madam Chair.

Okay, and we'll just do one final round? Okay. Um, I'd I'll double back on SBIR again, uh, wanna thank you, I'm glad we were able to get that over the finish line. Um, I did spend a considerable amount of time on reauthorizing the SBIR and STTR. And I really do appreciate the department moving quickly, uh, once those programs came back online in April. And we want to be able to transition the best of the best, uh, technologies to the warfighter as quickly as possible. That's long been my priority. And it's why we have now the creation of the strategic breakthrough awards so that we can rapidly scale the best with buy-in from our acquisition officials. And Dog and the broader autonomous systems enterprise rely heavily on these nontraditional commercial vendors for the software and the hardware, uh that will define the future war-fighting edge. And those are often the exact companies that really need a little additional help with their runway uh to get it to the war-fighter. So um Under Secretary Michael, what are we doing to ensure that the transition pathway from that prototype to actual production is actually functioning?

Yeah.

And then how can we better measure whether uh the SBIR authori reauthorization is meaningfully improving this process?

Yeah, I mean part of the insight that we had about unifying the innovation ecosystem was to have a place everywhere in the life cycle of a company, starting with sitter-sitter,

Mm-hmm.

all the way through a program of record. And that gives me some visibility, if something comes out of siber-sitter, then maybe it could get to DIU, the Defense Innovation Unit, then maybe APFIT, then the Office of St- uh, Strategic Capital. And then we could do test and evaluation so that it can leap to the services, without big gaps in time,

Mm-hmm.

as a valley of death.

Mm-hmm.

So, i now have visibility into all those sort of organizations and can ensure that i'm tracking a company all the way through good example of a company that's made it all the way through now is castilian they were a company now they're developing low-cost hypersonics less than half a million

mmm

dollars per missile relative to the fifty million dollars per missile we pay today and those are good stories so getting those stories out there and making sure that the services and each of these other funding sources that now have some oversight on can see and take those uh weapons forward are uh really, I think, gonna compress this notion of value in death and hopefully eliminate it.

Yeah, wonderful. Any other input um from the other members today, Doctor Wolfson?

Uh, yes please, I um just wanna say how grateful and thankful we are for the additional uh uh funding that you added in the new SIBER authorization act. We have actually had some of that similar authority through our Stratfi process, which is a SIBER funding. uh program of record funding and then venture capital funding um and we've had that going for a while and i just wanted to tell you a couple of the big successes that you know

great yeah

reinforces what honorable emil um honorable michael was saying um andrel was one of our first uh uh organizations that utilized uh stratfires uh they have since bought out another company that utilize stratfires called blue force technology and their current evaluation is about sixty one billion dollars um shield ai started as an AFRL Stratfy um that has also um helps develop our our autonomy for our combative collaborative combative um platforms and they're about twelve billion dollars now and then also Castilian was one of our AFRL uh Air Force uh Stratfys as well um and is doing great things so we are grateful for that capability and hope to continue to leverage it.

Thank you very much, Mister Tremper.

Uh yeah, in fact uh during your opening statements you alluded to relationships with industry and how do we foster those and how do we, how do we do things faster. Uh, one of the, the big Navy commitments is in that integration space and we look at our warfare centers distributed across the country as our seasoned experts on, how do you operationalize things, right? If I have a, a technology that shows up, it's a nontraditional, it's a s- it's an SBIR phase one, phase two, how do I then take that and make it operationally useful? How do I bring it into that phase three? How do I then measure success of my phase three investments versus my phase three non-SIBAR investments versus my phase one phase two. And what we're seeing is about a two to one investment return on investment, so two hundred percent increase in phase three over what we invested in phase one. That integration space is key though for things like Castillian. The ability of those vendors to get access to ranges, to get access to engineering expertise within the government through Mm-hmm. CRETAs through uh commercial services agreements allows them to understand the operational context of the technology they're going they're they're they're they're bringing their technology into. And now you can weave in the survivability and sustainability that then streamlines the scalability through acquisition. So we're seeing a lot of success in in those areas, and in fact to the lab infrastructure point. Uh, the commercial services agreements where uh, where a VC backed company is then creating an agreement with the government, allows them to pay to use the infrastructure of the government which then t- gets put back in to the government infrastructure. That's not government resources. and government infrastructure.

Mm-hmm.

That's VC resources being returned back into government infrastructure. So there's a lot of those part private uh public-private partnership arrangements that we're seeing a lot of benefit from SBIR's VC-backed uh in terms of offering our engineering expertise to outside vendors.

That's fantastic. Mister Manning, anything briefly?

Yes, uh, so we also recognize the importance of taking these small companies and getting them on a path to scale and prior to our most recent acquisition reform we launched a Cibber catalyst program which uh part of that was small business partnered with a larger integrator part uh connected with um at the time program executive office money so resources in all three everyone having skin in the game right at the beginning to go hey this is panning out how do we take it to scale uh more recently with Acroform we've aligned our Cibber program with army fuse connected to the pit so we anticipate having those small businesses going and iterating with soldiers of the eighteenth Airborne Corps. And when we see promising technology, the pit director connecting with our portfolio acquisition executive has to go how do we get this on a path to rapidly take it to scale?

Thank you. Okay, ranking member.

Thanks. Um, Under Secretary Michael, just to continue our line of conversation, you know, I I again in in a in a um s with a strong desire to win the AI race, it's hard to rationalize t- in my brain the two decisions of agreeing to sell some of our most sophisticated chips, NVIDIA chips, to the Chinese, and then secondly picking a fight with Anthropic. And I I know you've touched both of these, but um let me just ask on Anthropic as a follow-up and for others on the panel. Um so this fight erupts between Secretary Hegseth and Anthropic. It somehow can't be worked out, even though all of you up here use Anthropic in your classified systems. Um um there we've declared them a supply chain risk, and you've been told by I believe August to wean yourself off and extricate them from your systems. Um then they dropped the the detail that they've invented this thing called Mythos, which is gonna take us um light years ahead on cyber threats in particular. And it's such a leap that they are actually testing this commercial product like a weapon. They are giving it to private sector companies, and I understand they are giving it to you all to test and try before they put it out on the open market because they're so worried about the vulnerabilities it will expose, and the intended national security risks that come from that. Um, who knows the next thing they're going to invent, but the Department of Defense apparently won't have access to it. So for um uh uh separate from um Under Secretary Michael, for the the three others representing the services, are you currently using Anthropic in your classified systems today? And what is the cost um both in terms of capacity and money, since this is a budget hearing, in pulling that out of your systems by the August deadline if you intend to meet those? Let's start with Mister Manning.

Uh, Senator, uh, I'm I'm aware that we're using some, anthropic. We are in the process of assessing where all that is so that we could be in compliance with uh the direction from Department of War. Uh, I would have to take a question for the record as to what the cost would is associated with it.

Yeah, I'd appreciate that, and if that's gonna be the answer that all of you are gonna have to say, I'll hold you to it, cuz this is a budget hearing, you're asking us for money and decisions that the Secretary of Defense have made. is going to cost the American taxpayer and your budget's money. I'd like to know what that cost is. Um and um uh for the rest of you, are you currently using Anthropic in your classified systems? Um and are you gonna meet the time line to extricate it from your systems, and at what cost?

Uh yeah, the Navy is also in that one hundred and eighty day assessment period. I can't confirm uh current use of Anthropic in our classified systems but but we'll be happy to come back and provide that information.

Same answer. We're still within the the that hundred and eighty period. We are identifying it. Once we've identified where that is, we'll understand what that costs and what those challenges are. Okay, so I will um circle back with you all and expect to get that. I'd like to get that in a timely fashion. If you're in you're asking us for money, it's budget season, we are going to mark up the NDAA, then you can get me your answers on what you're gonna Um, and lastly, Under Secretary Michael, I would just say, um, what is complicated is that many of us, again, both sides of the aisle desperately want to win the the AI race against China. Um, but there's just, it just feels like story after story around corruption, um, um, every single week. And the story that came out this week is that the president bought or traded five million dollars worth of NVIDIA stock uh you know a week before he allowed the Department of Commerce to say that we could sell those chips those kinds of things speak to the perception of impropriety. I don't I can't prove guilt or innocence but the perception. And I know you got into it on the Hask side um with folks over there um I I can you just confirm for me separate from paperwork and when you sign different things, cuz we've had this discussion before have you made money in your portfolio on ai stocks since taking a job with the federal government where you're in charge of ai

uh i've recused myself from any company that i had a holding in

but have you made money by all accounts you've made significant amount of money it's public

i i have i made money because i had the vest of all the stocks that i might have had a conflict in so i had to like when you sell them you either make or you sell them

you sold all your stock from

so all my conflicting stock And if I'm for those that the the OGE said you can keep them, you just have to uh recuse, I I followed that faithfully.

So do you have any AI stock in your portfolio right now?

I don't.

Okay, thank you. You're back.

Thank you. I recognize Senator Kelly.

Thank you, Madam Chair. Mister Michael, under title ten US code section thirty two fifty two, the statute that your department invoked against Anthropic as a supply chain risk. It's defined as the risk that an adversary may, and this is a quote, sabotage, maliciously introduce unwanted function, or otherwise subvert, end quote, national security systems. That's the language that Congress chose, sabotage or subversion, language that implies the undermining of our national security. On an all-in all-in podcast in March, you explained this in your own words, saying, and this is another quote here, " What if this software went down, some guardrail kicked up, some refusal happened for the next fight like this one, and we left our people at risk?" So that was the danger that you described. A mid-mission failure that would put American troops' lives in danger. Now, here are the facts as we sit here today. Anthropic has been designated a supply ch- supply chain risk. And your secretary, Pete Hegseth, called the decision final. The president has ordered every federal agency to stop using their products, and yet, US service members have used Anthropics clawed in combat operations against Iran since the stri- strikes began on February twenty eighth. Your secretary's one hundred and eighty day phase-out has not expired. As of March twenty-fourth, your own Chief of Chief uh Information Officer, Kirsten Davies, testified to the Senate that Claude's use in US military operations in Iran was, quote, " active right now." So I've got a question. It's two parts. Part one, if you are correct, if Anthropics product is genuinely what section thirty-two fifty-two contemplates a national security supply chain risk capable of sabotage, or a mid-fight refusal that gets Americans killed, how does your department justify leaving that product in the hands of our troops in an active conflict one second longer?

Uh, when they were declared a supply chain risk, we stopped all updates permitted from uh that company, so we froze the code and the uh cloud provider that it was hosted on froze any access so that no change could be made in the code, so there was no risk of any mid-battle change in the operations of that

But that also means there was no risk in the code as it was on the day that you said

uh

it could be used against the troops in a way that could put them at risk.

No, Senator, because the uh before you came in, I listed the things that were against their terms of service. uh that for example you we're not allowed by their their terms of service that we live under today to do battlefield management applications target any military infrastructure of adversaries direct interceptions like of a drone uh or develop weapon systems and we just add the that could still be built into the code so we just avoiding the use cases

let me let me ask you a second question though if anthropics product is then it seems like it's what what you what they're using is safe enough now to remain in our soldiers' hands in a shooting war against a foreign adversary. How does the department justify branding an American company a sabotage and subversion risk under a statute that Congress wrote for foreign spies? Seems a little bit extreme. And one of these two things, you know, has to be wrong. And which is it?

Um, I The the sophistication which with which we use AI at the Department of War is very nascent right now. The way we use it is still very early days. What we're worried about with the terms of service that they had and their posture toward the department which when they questioned the Maduro raid and whether their software was used imp inappropriately gave us the sense that this was not a reliable partner to deal with, given their

Bec- because they asked questions about how their product was being used?

No, but Uh, that that

That means they're not reliable?

that in con in conjunction with their written terms of service which prevent the use cases that we would like to advance into, battlefield management, directing interceptions, developing weapons systems.

Yeah. Well, here's the thing, there seems to be a theme here. that anybody who questions this administration, questions the Secretary of Defense, uh, offers some other route, uh, some other idea that isn't consistent with the leadership of the department. Uh, you know, they're branded, in this case, this company, branded as subversion risk and essentially fired because they started trying to have a discussion about hey, what's uh, you know, what's the best use of this or why did you do this this way or that way. I think, you know, my - my view on this after, you know, a lot of - lot of experience in the military and at other government agencies, if you're not asking the hard questions, especially uphill, you're not getting the best out of your people, and you're not getting the best decisions. And my concern here is you - is you branded a company uh, in a way that's harmful for, to, for them and their employees, you know, thousands of people that, that work there. But you also may have ultimately put our service members at risk, if they are in fact the best company to pr- provide these services. They seem to be. Thank you, Madam Chair.

C- can I c- can I get a question? Uh, yes, and you may, because um, uh, uh, Secretary Michael had listed now the, uh, partners that we do have that will be filling this gap, but yes, I'd like you to respond.

Yes,

Thank you.

uh, Senator, every other major AI company, every single one agreed to our terms. So this wasn't about questioning whether we were, um, you know, they dared to question us. This is, we wanted to, to use their products for all lawful use cases. And Microsoft agreed to that, OpenAI agreed to that, NVIDIA agreed to that, Google agreed to that. So every other company agreed.

How many of those were after Claude essentially got fired?

They were all after. They they were all after because we didn't we

All of them?

only had one provider before we got in this seat.

Right.

And when when we looked at the contract to that provider, we said we can't grow with that provider.

Think there's any chance they saw what happened to Claude, saw an opportunity, and then said, well, I know what I have to agree to.

Absolutely not. Um, maybe OpenAI, I'll I'll I'll retract and say maybe one company that sort of tried to slide in, one of their but Google has been a long-time partner of the department, Microsoft, Nvidia, real, real big companies with proper corporate governance, went through their legal teams and agreed to our terms of all lawful use cases where Anthropic would not. So that should say something that our terms weren't unreasonable, we weren't trying to be unreasonable vis-a-vis any one partner for any specific reason that they didn't demonstrate themselves.

Who do you think is best equipped to provide these services today?

It depends on the application, I'll give you some examples as to why. Um, Google for example has a lot of video imagery because of their YouTube and Nest cams and therefore from a robotic standpoint they may provide better capability from an autonomous weapon standpoint um NVIDIA has an open source product which is gonna be a crowd sourced capability that can compete with the Chinese open source models which are currently infiltrating a lot of the US companies. So each of them have different capabilities, which is frankly why independent of Anthropic, I would have ensured that we had multiple providers, cuz the leaderboards change on these providers every few months.

Mm.

It wouldn't surprise me if next month you have, opening, I have a better cyber model than Mythos. So for us to be responsible to the war fighter, I've gotta have all the options possible as these models change in advance in their role.

So what do you do about Mythos now?

Yeah.

Cuz that's clogged.

I think we'll start wrapping up, yes. Thank you.

Yeah, I I think mythos is the first cyber weapon model. Hope and I will have a cyber weapon model. Jim and I will all this calendar year. And then next year we'll be worrying about bile and chem ma- models as they start to train them with science and physics and equations and and math. And we're going to have an ongoing discussion probably and hopefully with Congress about how we govern the rollout of these things that have good effects. Maybe they can cure diseases. But then they have bad effects when in uh the hands of bad actors, and that's something that we're gonna have to deal with as a nation.

Mm-hmm. Thank you. And I think uh this entire lesson is that I believe um when the department was negotiating those contracts with Anthropic, we should have had the foresight or the department should have had the foresight to really question uh what was in those contracts unfortunately that was not done, and we're in the situation we are today. So I really appreciate all of our witnesses here today. This is an important discussion as we move forward with uh really ever-evolving technology. Uh so, really appreciate your presence and your testimony in front of the subcommittee on emerging threats and capabilities today. And with that, uh we call this hearing to a close.

Right.