Hearings to examine China's ongoing theft of U.S. innovation.

Good morning. Welcome everybody. This hearing will come to order. Senator Grassley regrets not being able to attend uh in person today. He'll be entering or I'll be entering his uh opening remarks into the record, along with uh uh statements from uh the three other members. Um, today's hearing will explore how foreign adversaries specifically the Communist Party of China, threaten US economic and innovation dominance through rampant intellectual property theft. The economic loss from China's IP theft is estimated to be between four hundred and six hundred billion dollars a year, the equivalent of costing the average American taxpayer payers dollar, thousands of dollars every year. China's overall goal is to undermine US leadership to assume the title of world's innovation leader. We cannot allow this to happen, ladies and gentlemen, and while the theft is a major concern, we also recognize the concern of China evolving from an imitator to an innovator. Uh, before I um uh yabbled down, I spoke with two of the witnesses to talk about the number of committee meetings that we've had in the intellectual property subcommittee, just documenting just theft, blatant theft, blatant uh, counterfeiting, uh, under our s- nose, and I think it's time for the President, the administration and Congress to take action to hold China accountable if they themselves will not take action. The US must overcome its historical and ideological views that China is unable to innovate. Innovation and creativity are critical to the economic and global competitiveness of our great country and to our national security. IP intensive industries generate nearly half of the US gross domestic product and employ over set forty-seven million workers. Therefore, we must do everything to ensure that strong, clear, reliable and predictable IP rights remain central to this country. Make no mistake, we're in a race with China. But unlike China, we are not in a race to the bottom. And uh, I I think again, I'm glad to see that Chair Grassley has convened a a full committee on this subject. We've been talking about it for years, but we've only made incremental progress. And now I think it is time to educate the broader members, uh, the base of members on this issue so that we can work together on a bipartisan basis to bring about real change. And now I turn to Senator Durbin, Ranking Member Durbin, for opening comments.

Thanks, Senator Tillis, and I also want to thank Senator Grassley for calling this important hearing. i spoke to him a couple of days ago he is doing great expects to be back next week uh and i wish him well i also wanna thank senator tillis for serving as chairman today thank you to the witnesses as well the us the united states has been the world's leading innovator from the light bulb and telephone to the semiconductor and gene editing americans have been at the forefront of many of the world's greatest breakthroughs but us leadership in this area is never guaranteed Over the last fifteen months, President Trump has repeatedly taken actions that harm our innovation ecosystem. Let me be specific. He's attempted to cancel grants, fire scientists, and freeze billions of dollars of the National Institutes of Health and the Ni- National Science Foundation funding. He sought to freeze res- university research funding for our top universities, funding that helps develop technologies of tomorrow and allows us to maintain our lead in innovation. That's not all. He has imposed on-again, off-again tariffs with little regard to whether they target our allies or adversaries. Excuse me. As a result, our nation's reputation has been shaken, making it more difficult to build the coalitions and engage in the international organizations necessary to collectively counter China's growing global influence. China is taking advantage of American chaos during whatever it takes to overtake our position as global leader in innovation. Several decades ago, China launched an all of country approach to use all available means to acquire foreign technology to support its efforts to become a global leader in technological innovation. In twenty fifteen, China announced its Made in China twenty twenty five initiative. calling for Chinese technology and advanced manufacturing to catch up to and replace foreign technology. On a level playing field, I like our chances of being the first to develop next-generation batteries, quantum computers, and new cures for diseases. But too often, we aren't on a level playing field. And the current administration's efforts sadly, in many respects, are tilting us toward China. Why in the world are we taking the most sophisticated computer chips that we have developed and sending them to China doesn't make sense to me. China has engaged in economic espionage and IP threat theft to steal work that may have taken many years to develop at at an estimated cost of between two hundred and twenty-five and six hundred billion dollars every single year. More to the US economy. This effect affects both the economy and the jobs of hard-working Americans. US companies have less incentive to invest in R and D and other jobs if they think China's gonna steal their patents copyrights trademarks or trade secrets. Take the case of Lin Wei Ding. While employed at Google, Ding used stolen AI related trade secrets to start his own AI company in China. He was convicted of theft of trade secrets, but the Chinese government retains those secrets and can use them to their benefit. Chinese tactics are not limited to economic espionage. It also acquires IP through agreements that require companies that want to do business in China to partner with and transfer technology to Chinese companies. Once the technology has been transferred, the Chinese partner can transfer it to the Chinese government. This happens far too often. A few years ago I learned of the story of Enios, a company which has an R and D facility in my state, that developed a chemical which is a key building block in hundreds of applications across the medical, aerospace, environmental and military sectors. Senepec, a Chinese oil refiner, stole INEOS trade secrets and began manufacturing this chemical. INEOS sued in the Chinese court system. The companies reached a settlement under which they agreed to jointly produce the chemical. Soon after, engineers working on the joint venture left to create a competing company in China using N EOS technology. N EOS experience is not unique, and these stories all seem to result in China taking American companies' IP to create their own domestic industry. It should be said, and this is important, that while today's hearing focuses on IP threads from China, the government and its companies, China is not the only country engaged in or hosting entities engaged in IP theft. Other countries like Russia have done the same. We must ensure our efforts to address the problem are aimed at the real bad actors, and we don't simply target people of Chinese descent, which is too often the reaction in these situations. I look forward to discussing how best to address these issues, and how we can protect American innovators and technology they create.

Now turn to introducing today's witnesses. Uh first we have Tom Lyons. Mister Lyons is a former senior officer with CIA. He has more than twenty years of experience in the US government and the private sector, focused on counterterrorism and China related issues. In twenty twenty one he helped found the Twenty Four Thirty Group, a non-profit, non-partisan research institution that focuses on state sponsored espionage against Western businesses. Thank you for joining us, Mr. Lance. Next we have Helen Toner. Miss Toner is the Interim Executive Director of Georgetown Center for Security and Emerging Technology. She previously worked at Open Philanthropy, uh, uh, Open Philanthropy, Philanthropy, where she advised policymakers on artificial intelligence, policy and strategy. She also previously lived in Beijing, where she studied Chinese AI ecosystems. Finally, we have Mister Cohen. Welcome, Mister Cohen. It's good to see you again. He is currently the Senior Technology Fellow at the Asia Society of Northern California, and an Edison Fellow at the University of Akron School of Law. He also serves as a Non-Resident Scholar at the University of California and the National Bureau of Asian Research. He previously served in office in the Office of International Relations at the US Patent and Trademark Office. His career has included more than thirty years of experience on the issue of intellectual property rights in China. On behalf of the committee, I want to welcome all three of you here today. We look forward to hearing your testimony on this important topic. As a tradition of this committee, I ask the witnesses to rise and uh raise their right hands. Do you swear or affirm that the testimony you're about to give before this committee will be the truth the whole truth and nothing but the truth, so help you God? Thank you. I'll let the record reflect that the witnesses have answered in the affirmative. Uh, Mister Lyons, we will begin with you. Senator Tillis.

Ranking member Durbin, committee members, good morning. I spent a career at CIA focused on foreign adversary capabilities and intentions, not intellectual property theft. But you cannot work in the intelligence community without seeing evidence of China's economic espionage throughout the United States. Front companies operating freely, talent programs, strip mining our companies, and core infrastructure built on stolen American secrets. The magnitude of the PRC's effort we saw was staggering. The response, however, from both government and private sector was not. This disconnect is why I left government. It's why I helped found the twenty four thirty group, and it's why I'm here today. Our team has a mission that we could not have pursued from inside government. We partner with and help US companies defend their innovation from theft. And what we've seen has been alarming. American firms are not competing against Chinese rivals in any normal sense. They are competing against the largest intelligence apparatus in the world. One whose mission includes putting American companies out of business. This is not GM versus Ford. This is a US start-up versus the resources of the PLA. The issue is that our system was never designed for this. Our commercial sector and legal system assumes good faith, self-interested, profit-oriented companies, but none of those assumptions hold up when the other side compels theft sponsors copycat products at subsidized costs, and has little focus on profit. American companies are defending themselves against the ministry of state security and approaching it like a compliance issue. The consequence of this mismatch will be the wholesale transfer of American industry to our adversary. Look at what they've already achieved. To highlight this disconnect, consider Lin Wei-ding, who stole Google's latest generation TPU designs, technology at the center of a trillion dollar industry. What is most striking about Ding is not the theft itself, it's rather that he advertised the theft as a comp- as his company's differentiator when he was back in the PRC. The theft was used like a credential. And it's important to note that while Ding was caught, that technology is gone and we will only see the consequence of that in several years through a low-cost copycat. What about American Superconductor? A Massachusetts company that saw the theft of its software from a PRC state-owned company. In response to this multi-billion dollar theft, the PRC company paid a fifty-nine million dollar fine. And it got to to keep the stolen software and continue to sell it. Meanwhile, American Superconductor had to fund its own litigation across multiple jurisdiction for years, hemorrhaging revenue, fighting to survive, all the while competing against its own products. How about Apple's autonomous vehicle program, which was hit by three separate major PRC theft cases? DOJ went after the individuals, but not the companies that commissioned the theft or the system that incentivized the behavior. And today we know Apple has no autonomous vehicle program, yet the PRC is the fastest growing on the planet. In the PRC system, while an individual may steal the technology, they're doing it on behalf of a system that rewards the behavior. It's prestigious, it's lucrative, and in some cases obligatory. In the US we are not addressing the infrastructure that supports nor do we address the companies that benefit from the theft, the occasional DOJ conviction does not unsteal the technology. We must do more to get in front of these crimes. For example, we should increase the deterrence mechanisms by making the consequences of theft devastating for those involved increase the penalties under the Economic Espionage Act as it stands they serve as a small tax, relative to the benefit received. We must ensure that the beneficiaries of trade secret theft, not just the individuals, face consequences. Consider creating a key TAM enforcement model, so plaintiff companies can both share in the financial recovery and do not have to share in the litigation costs. Pair this with a whistleblower bounty program, modeled off of Dodd-Frank, providing financial or immigration benefits to those who step up to protect our companies. We also should mandate that every federal contractor undergo foreign influence background checks. And we must give our company's federal authority to conduct

Mm-hmm.

these same checks on employees. Currently in some states, it's prevented. Let's build a proactive incentive structure by offering tax credits to corporations that invest in counter-intelligence programs. Detection is everything. And companies cannot report what they don't discover. Likewise, we propose a federally backed insurance program for trade secret theft by state actors so that companies have a financial incentive to report rather than remain silent. If a foreign military were conducting operations on American soil, we would not ask our companies to fund their own defense, yet that is the current reality. The PRC is targeting our infrastructure, our companies, our civil society on our soil. And we've, we've outsourced this issue to IT or HR departments, or frankly, to CEOs whose primary objective is next quarter's earnings. Our issue is a national security failure. But this committee can change that. Thank you and I look forward to your questions.

Miss Toner, before you begin, I can assure you I know how to pronounce philanthropy. May begin.

With staff assistance.

I appreciate it, Senator Tillis, uh, ranking member Durbin and and senators, uh, thank you for the opportunity to testify today. My name is Helen Toner. I lead the Center for Security and Emerging Technology at Georgetown, where for the past seven years my work has focused on AI and national security, US-China competition over AI, and AI safety and security issues more broadly. Accordingly, my testimony focuses on issues related to AI, both given that it's where my experience lies and given the vital strategic importance of AI today. Right now, the AI-related IP threat that is receiving most air time is known as distillation. This is a technique whereby a more advanced AI model is used to improve a less advanced AI model. Distillation is a real threat to US AI competitiveness. Chinese AI companies are distilling from American models at large scale, and this is something we should work to understand and prevent. But distillation is far from the only threat to US IP or to US competitiveness in AI, and it would be a mistake to focus myopically on tackling distillation at the expense of other threats. As I outline in more detail in my written testimony, a simple way to think about types of IP that AI companies must protect includes four elements, AI models themselves, algorithmic secrets, training data, and AI hardware. There have been multiple cases of these different categories of IP being stolen. In twenty twenty three, hackers breached OpenAI's internal communication systems and extracted extensive internal documentation including algorithmic secrets. Earlier this year, as already mentioned a couple of times, a Chinese national formerly employed at Google, was convicted of systematically exfiltrating trade secrets from Google, including large amounts of information related to the company's state-of-the-art AI hardware. As for training data, companies including OpenAI, Anthropic and Google have reported large-scale distillation campaigns by Chinese actors on their platforms which essentially amount to extracting training data from US models. There are also more direct routes to acquire training data, including what hackers did to provider uh company, US company named Mercore two weeks ago, where they simply hacked into Mercore's servers to steal proprietary data that they can use to train their own models. It's also possible for Chinese AI developers to simply buy discounted data and training environments a few months after US companies are done with them as part of their fast follow approach. All of that to say there is no shortage of ways in which US innovation in AI is vulnerable to foreign technology transfer efforts. So what should we do about it? I have more detail in my written testimony, but briefly some highlights. One recommendation at the strategic level is to make sure that any effort to tackle our concerns about distillation is done in ways that support rather than undermining other elements of US IP protection and AI competitiveness. So that means don't divert resources away from threats like preventing adversaries from misusing AI more broadly or from stealing other types of AI IP. And don't pursue approaches that would undermine the open and international AI research ecosystem that has been the fundament of US success in AI. More tactically, four very specific recommendations. One is to expand funding for the AI security center within NSA, and the center for AI standards and innovation within the department of commerce. These are two programs that facilitate information sharing about security threats between government and industry. Two, in service of this information sharing, DOJ should expedite security clearance processing for relevant AI company employees, which is often taking a long time and delaying valuable information sharing. Three, DOJ should provide clear guidance to AI companies on the extent to which those companies may coordinate on defending against AI distillation and other misuses of their models without being subject to antitrust infringement. It's been publicly reported that top AI companies like Google, OpenAI, Anthropic are undertaking some limited steps to share information with each other, but in private conversations I hear that they feel their hands are tied by legal ambiguity around antitrust concerns here, which DOJ could easily clarify. Fourth, in uh in service of uh better information sharing in the AI space more generally, I commend Chair Grassley's AI whistle-blower protection act, which can help ensure that when employees of these powerful companies uh see something wrong, they're able to speak up about it. Thank you and I look forward to your questions.

Thank you, Professor Cohen.

Thank you, Acting Chairman Tillis, uh, Ranking Member Durbin, uh, members of the Judiciary Committee, it's my pleasure to testify before you today. Uh, my name is Mark Cohen. I've worked on US-China intellectual property issues, core IP issues for over thirty years. Uh, I'm affiliated with the Asian Society of Northern California, University of Akron, other institutions, But the opinions expressed today are solely my own. I'm gonna cover three topics. First, but to be a little bit swimming upstream in this discussion, the IP theft narrative has, in many respects, misguided us. There's certainly IP theft out there, trade secret theft out there, uh, a forced technology transfer, uh, but our policy response in general has emphasized public enforcement, that is, criminal remedies, uh, and headline cases, rather than systemic issues which And IP is primarily based on civil enforcement. And in China we're talking about five hundred thousand civil cases per year, of which foreigners are only a small component. In many respects, by focusing on criminal remedies for over two decades, we've helped enhance the role of the state. We need to look at IP as a private right, which it is what it is. We do not track how foreign companies actually use China's IP system, including its civil system and actual outcomes. We have underemphasized the importance of transparency, which has been declining in China, and makes the phase one trade agreement largely unverifiable. So we face a series of paradoxes, large claim losses, but only selective publishing uh publishing of cases, uh extensive policy activity, but limited verifiable participation in the outcomes, significant and tragic anecdotal events, but agreements that are impossible to believe. fully verify. We are making policy in the dark, without a clear understanding of how China's system is actually used, where it succeeds, and it does in some places, and where appropriate corporate or government resources need to be allocated. We have also not sufficiently focused on strengthening the institutions that enable private rights to be effectively enforced in China. And this, by the way, is not new. The USSR had one of the ten largest patent offices in the world. for decades. So China being number one is actually not new in the IP context. We undermine in some respects, uh, the patent system in the United States, in our own country, by placing our inventors at risk, by weakening eligibility, uh, limiting injunctive relief, and under-emphasizing the role of civil enforcement. Second, there is good news and bad news about China's IP system. The good news is China believes in IP. The bad news is China believes in IP. Uh, China has scale, speed, lower costs, broad injunctive relief, and comprehensive industrial policies to advance its IP system. But it is state driven, and IP is used as a tool of industrial policy, with subsidies, metrics, administrative guidance, and a non-independent judiciary. We are competing with a peer that has a lar- is the largest filer of international patents, It has the largest patent and trademark office, it has the largest civil and criminal IP docket, and it uses these systems to advance national objectives. It may also shortly overtake the United States to become an international norm setter in intellectual property, which I think would be tragic. China is also moving to shape policy in emerging areas like AI and data governance, including, for example, treating data in ways that resemble a form of intellectual property. Chinese policymakers also react quickly to address IP challenges and new and emerging technologies. As we consider patent reform and the implementation of the phase one agreement, we should recognize that China is reinforcing key features of its system through a system that often relies on public law remedies. We need to smartly to collaborate with China as well to address common challenges and to maintain one of our great the great strengths of our system, which is relative openness uh uh to outsiders to participate. Even our court systems compete. Our so-called rocket docket, the ITC in the Eastern District of Virginia, is slow in compared to China, which has become a pacing competitor in cross-border IP disputes. We need to ensure that our private rights system continues to be a world leader and model for the world. Third, we should not be surprised by China's advances, and we are. China is not only copying, it is increasingly innovating. Recently, for example, we've heard a great deal about China's rising leadership in the biopharmaceutical sector. Many of these advances, if not most, could have been anticipated. Uh, decades ago, uh, the Congressional Office of Technology's assessment warned of this. Chairman Grassley, who I regret is not here, I wish him a speedy recovery, was on the board of OTA. I have brought here a copy of OTA's July nineteen eighty-seven report, which anticipated major advances in China's use of uh development of dual use technology, and and a general narrowing of the technology gap by two thousand seventeen. It was prescient, OTA is gone, we did not maintain that institutional capacity, we need to rebuild it. I believe USPTO can be very helpful in that regard, but most importantly we need to recognize that we need to maintain effective competition and continuing leadership in IP and that burden is on us. Thank you, uh, members of the committee.

Thank you, Mister Cohen, and, uh, appreciate you being back before this committee. We've appreciated the, uh, contribution you've made in the intellectual property subcommittee. I'd like to start by talking about something we have discussed at the, uh, subcommittee level, and that's the Patent Eligibility Restoration Act. Um, the, you know, we've, uh, uh, we're we're hopeful that we're gonna have a full committee hearing on it. uh before this congress ends but could you talk about that bill uh uh with respect to china uh how how you think that that may be helpful or not in your opinion

well let me state first of all that we've had a lot of judicially made doctrines that uh were not uh legislated that have limited the scope of patent eligible subject matter in emerging technologies

mmm

now the comparison to those cases uh has been uh

yes

China has amended its patent examination guidelines about eighteen times since nineteen eighty five. Uh, and it's amended its patent law about five times since about two thousand one. So what you have in the institutional structure in China is a patent office and, more importantly, and an industrial policy network that leans in heavily on where the emerging technologies are, develops that information from five-year plans which made in China twenty twenty five, with all due respect is only one and the least granular uh uh you have them uh by locality at the national level uh and often voluntary ones by industry groups and those become the guidelines going forward from the patent office what we've done is had a very static system that is slow to legislate slow to the to react while china has granted patents to the same areas so what are the consequences if you're an inventor here in an area in genomics and business software and many other areas an AI, uh, your patent will be granted in China, perhaps the EU as well, will not be granted in the US, and there are many instances of that, and that drives the innovation overseas.

Andy, with some of the pushback that we've gotten on Para here, um, we're we're not trying to accelerate someone who ultimately gets a patent, we're just trying to open the buy box on what may be patentable. Do you agree with that?

I agree with that completely, yeah.

Do d- do you see any downside to that?

Well, some people believe that patents shouldn't be protected at all. Uh, uh, and, but, you know, we have to be really careful about that because there's some areas of technology that are less patent dependent. Uh, uh, there's some areas of technology where you can get patents and share them. And actually the patent system doesn't require that you get paid, although most people expect that. Uh, but they're obviously open source in other mechanisms. But in biopharmaceuticals, it's absolutely critical to get a patent. In many areas of high tech, it's absolutely critical.

Yeah.

Uh, and the more we weaken the system, the less we're capable of playing as an effective partner in, uh, setting global standards that are patent-dependent, and in many, many other areas. We talk a lot about these high profile national security areas, but there are many other areas, uh, in addition to that.

Thank you. Um, Mister Lyons or Miss Toner, are you all familiar with the BEARD uh act that we put together? The Bad Electronic Art Recording Distribute Recording Distributors Act. It's basically, I I didn't I didn't wanna uh throw a question at you if you hadn't had a chance to look at it, but what we're really trying to do is align US law with what I think are emerging standards uh uh in Europe. I'm not one to gold plate uh European policy in some cases. I think they got it wrong with GDPR and a number of other things. But uh what are your thoughts on joining other countries around the world that are identifying and protecting consumers from these piracy web sites, if you're if you have any opinion on that? And then, Mr. Cohen, I I know you probably do. I'll come to you last.

Senator, thank you. I'm I'm not familiar with the specific act that you're referencing. Uh, I do think this is an international effort, and I do think that, uh

You can get your mike a little bit closer to your

I I do think this is an international effort. And we, in order for us to

Yeah.

In order for us to move the needle on this issue, we must cooperate with allies. Uh, during the Cold War we had COCOM, and COCOM had a lot of leaks, a lot of holes in it. However, we need to i think a the match act was proposed as well which is an allied export

mmm

uh prevention scenario and i think that uh the international front is is um important here

miss toner

i would agree with that i think certainly considering the multilateral aspects of anything we're trying to do here is is valuable

yeah one thing you mentioned about distillation i think probably the the first report at one of that was deep-seek and uh going after chatgpt which chatgpt just mentioned was a violation of its terms of use what more do we have to do to counter distillation now the concern i have with that like after uh uh after the distillation of the chatgpt l l m um is that that becomes an enabler for rampant intellectual property theft i p theft uh at and so those that that's what what what what should the industry do to counter that versus say don't do that i mean i i i get technologically, the challenges they have, but what are the trends there for the industry to protect itself from Chinese, uh, uh, initiatives like the deep sea distillation?

Yeah, I think the way I'd suggest understanding this is as one type of AI misuse. So different ways that, um, adversaries, bad actors try to misuse AI models. So other examples would include using AI models to plan cyber attacks or carry out cyber attacks or otherwise access kind of illegal information. Um, I think the, the best lever and the simplest lever that is available right now is clarifying antitrust guidance here. Um, it's a complicated technical problem for the companies to figure out what kind of filters do they put on, how do they detect this, how do they look at coordinated activity across different fronts. Um, and they are pretty wary of antitrust enforcement here if they, if they talk too much about what they're trying. Um, so I think there could be meaningful improvements, um, if they're just given a little bit more, more clarity on that.

Thank you. Senator Durbin.

Thanks, Mr. Chairman. Um, Karl Marx, who I don't often quote, said a hundred years ago, you can count on a capitalist to sell you the rope you plan on hanging him with. I'd like to ask you about this Nvidia chip sale to China and in the context of American innovation and competition for innovation, how this makes sense that we would take sophisticated computer chips and decide that China can buy them, which was the president's announcement, so long as we get twenty five percent on the deal. Are we selling our souls here by giving away this kind of hard-fought innovation?

I will say I think, uh, controlling access to chips and to computing power is really the best lever that we have in AI competition with China. Um, there's a lot of factors that go into AI competitiveness, and a lot of them are very difficult to control. Um, and so the fact that we are actually able to restrict the amount of computing power that resides in China and that Chinese companies have access to, is a really important lever. So I, yes, I think it is a mistake. I think it is also really important, sometimes in these discussions, um, the numbers get lost and the numbers of chips we're selling and the the level of how advanced those chips are matters a lot. And so I'm encouraged that despite, um, a lot of the back and forth on this policy issue, um, in fact, the licenses that have been approved, as I understand it, to date have been relatively small numbers of chips. Um, I think if we in principle approve sales, but in then in practice, the numbers are low, then, you know, we've sold them a maybe a uh a thinner uh less useful rope, um, uh, to to pick up on your analogy.

Anyone else care to comment on this? Professor Cohen? You need turn on your

Turn on the mike.

Oh, there you go. Sorry, a personal comment and an observation. I first heard about NVIDIA in two thousand eleven, uh, in an in a magazine called Electronics Intellectual Property. is where i first heard the term gpu uh and it was pretty clear where china was headed back then uh our inability or uh uh um lack of use of future oriented technology assessments which were pioneered by ota but and are widely used in europe there are even patents on ota approaches is really critical here in assessing what happens when you deny uh the profit stream for advanced technologies whether the us can remain uh, the leader in that area, whether or whether we'll be hampered from it. So, it sounds highly problematic to me to sell these chips to China, uh, but it may be more problematic not to sell the chips to China. And I think we really need to understand that much better.

Thank you. Uh, I also would like to address this AI situation, uh, the current state of research, and profess at the beginning, I am a liberal arts lawyer, So you lose me in a hurry when you get deep into technology. But as I understand it, we are basically inviting the private sector in America to get into the race for the AI uh achievements. Uh, and many of them are. They're, I think, four major companies in competition i- in our our national space for that possibility. When I look back in history, whether it's the A bomb or the man on the moon, uh, it used to be a partnership. between the government and the private sector. Can you describe the current situation in terms of the government's role and our control over the research and what comes from it? Miss Toner?

Yes, I think it's truly unprecedented, the mismatch between how strategically important this technology is, and how little involvement government has in developing it. Um, uh, and I think that is n- uh, most likely not a good thing. Uh, I sat before a subcommittee of of this committee, uh, a year and a half ago, and was was asked to testify about, um, what it is that, uh, you know, we in policy conversations in Washington, um, don't understand about insider conversations about AI. And what I testified in in September of twenty twenty four is that what we don't understand is how deadly serious they are about building machines that will, uh, outperform humans at everything. Um, and how deadly serious they are that they don't know if they'll be able to control the machines that they Um, I think it sounds like science fiction when you say it like that, but it is truly what some of the wealthiest and most powerful uh are pouring hundreds of billions of dollars into doing, and so we need to take that seriously.

Mr. Chairman, could I ask one more question?

Absolutely.

Mister Lyons, I wanna ask you a question that comes before me on a regular basis. Universities, major universities in my state are lamenting the fact that foreign students are uh choosing not to attend college in our country. A lot of it has to do with all the anti-immigrant conversation and the experience many of these immigrant and foreign students have had when they came to the United States. So I would like to ask you, with your background from a security viewpoint, do you believe we can bring, invite students to come study in the United States in some of these critical areas, without compromising our national security?

Thank you, Senator. Uh, I think that's a it's a complex issue. I think the design

If you could pull the mike closer to you.

The the design problem with um that we're facing is is how do we distinguish between those who are a resource and those who are a risk uh to our country? And I would say that as a country of immigrants, we need to always support uh immigration towards the side of those who are resources. Right now there's um very little checks that go into place when students and or Um, even H one B visa holders are arriving to the United States, and I think that, um, there is a spectrum of opportunity that we have that goes from increasing the amount of uh research that is done on individuals coming into the country, to an outright ban. I think the outright ban is something that's a bit off the table for now, um, and it should be. Even during the Soviet era, um, we had uh, we were pro-immigration. we should remain in that.

Thank you.

Senator Cornyn.

Mr. Toner, you are no doubt aware that uh CISA, the Cybersecurity and Infrastructure Security Agency, is a component of the Department of Homeland Security. And because of uh uh the uh impasse over funding for the Department of Homeland Security, they've now uh laid off uh or furloughed about sixty percent of their staff. Isn't CISA the primary federal government agency that's supposed to focus on this very cyber security threats that we face from China?

It is, and it's notable that this is happening at the same time as we have an American company, a leading American AI company, um, that has developed an AI system that can, uh, find vulnerabilities in every major operating system and, uh, and exploit those vulnerabilities. And that that company, Anthropic, is, uh, working to partner with uh, cyber defenders, um, including in both corporations and and governments. Um, and CISA would be a a really natural partner for them to make sure that we can defend and patch all the vulnerabilities that that AI system is able to uh identify, before it is released more widely. And it, I think it is, um, a real problem that the US government right now is not positioned to, um, to partner with them in that way.

I agree with you, and I think uh, it demonstrates the monumental stupidity of the failure to fund the department of homeland security including all of its components, but uh including uh including CISA. Mister Lyons, um the national security agency is our NSA is our premier cyber security uh agency of the federal government, is it not?

It is.

And uh the head of the NSA typically is dual-hatted um um cyber comm is the military um component there that is responsible for offensive uh cyber security efforts, correct?

Correct.

And um so it strikes me that we've primarily been playing defense when it comes to cyber theft um which is one of the specialties that uh China has. They just keep stealing our stuff and because we are not going on offense enough uh to raise the cost of those sorts of attacks and theft, uh they just keep coming. Is that a fair statement? Do you agree with that?

Uh, I think I would differentiate between the, a lot of the cyber theft. There is truth in what you say when it comes to the malware that's been placed on our core infrastructure. However, there's also a great deal of effort placed on stealing um, economic secrets, trade secrets and so forth at the PLA or the MSS. uh, conduct. And, and I don't think that's the domain of the NSA.

Professor Cohen, um, you know, we uh, obviously contrasted with uh, the People's Republic of China, we're an open society. And uh, it would, and that, they prey on our, the fact that we are an open society and they're not. Um. One of the areas where uh, we are an open society is in our, in our courts. and on litigation, uh, where people who have a grievance, uh, are entitled to go to court and seek redress, uh, for those grievances. But one of the things that's sort of come up on, in fairly recent years, certainly since I quit practicing law, has been this, uh, third-party financing of litigation, uh, which is relatively opaque. And what I'm concerned about is the potential and maybe the reality of, um, adversary nations, uh contributing to the financing of litigation in order to steal our intellectual property either through the discovery process or otherwise. Is that, is that a concern?

I, uh, you know, I, I think, uh, there's been little to document, um, uh, uh, non-practicing entities, uh, being funded by the Chinese government or surrogates. Uh, I do agree though, that uh we could do a better job across the board in encouraging courts, regulatory agencies, the pto the sec uh to disclose real parties and interest whether they're

so a little transparency would go a long way

and and and uh there's a professor at stanford uh who has written extensively on the uh labyrinthian links between the party state owned enterprises partial ownership et cetera. This is gonna vary by regulatory agency, but for example, CFIUS needs to look at who will control the technology, uh, when a foreign party is going to acquire a US company. So that would be critical. For the USPTO, it's who is gonna own the patents, so who is gonna be in charge of the patents, just like another technological product. And we saw this recently in a patent office proceeding which looked extensively at the real party in interest.

Five minutes goes fast. Thank you.

Senator Whitehouse.

Thanks very much, uh, Chairman Tillis. Um, I've got a little history here, uh, much of it working with Senator Graham. Uh, back in two thousand thirteen we released draft legislation to strengthen criminal prosecution of trade secret theft and economic espionage, um, particularly by means of state-sponsored cyber hacking we obviously had china in mind when we did that uh the next year we had a hearing in my subcommittee on crime and terror on that pattern of theft of intellectual property um ultimately congress passed the defend trade secrets act which created a civil cause of action for trade secret theft it included uh senator grahams and my provision making trade secret theft and economic espionage RICO predicates for uh racketeering investigations. And um here we are still concerned about what years ago I'd described as the greatest theft in world history. And by that I was referring to Chinese cyber intellectual property theft from the united states our first uh cyber command chief keith alexander borrowed that phrase and used it widely so what do we best need to do to empower um companies that have been hacked to seek redress uh one of the ideas we kicked around was to um, consider a government license to certain companies that would allow them, with inappropriate guardrails, to hack back at a cyber intruder. You obviously don't want a complete free-for-all. Um, but neither do you necessarily want the US government to become a choke point so that very little can be done when a company sees itself under that kind of attack. If you could just give me your best idea or two, I'd appreciate it, and we'll start with you, Mister Lyons.

Thank you, Senator. I think the biggest discrepancy in the private sector today is the fact that there's no incentive to pursue these cases oftentimes. Um, and while cyber is a portion of it, this is also a human game, very much.