Protecting Our Edge: Trade Secrets and the Global AI Arms Race

2025-05-07

Loading video...

Summary

The meeting focused on the critical issue of protecting trade secrets within the global artificial intelligence (AI) arms race, highlighting the inherent tension between promoting transparency and safeguarding innovation vital for US leadership[ 00:32:11-00:32:27 ] . Participants discussed the necessity of strong protections for proprietary algorithms and datasets, which are considered the "secret sauce" driving American AI advancements[ 00:32:45-00:33:00 ] . The discussion underscored the potential risks of intellectual property (IP) theft and the strategic implications for national security and economic competitiveness[ 00:33:08-00:33:58 ] .

Themes

Protecting US AI Innovation and Trade Secrets

Speakers emphasized that the success of the US AI industry relies heavily on safeguarding trade secrets, which are seen as critical for maintaining a competitive edge and driving economic prosperity. The unique adaptive nature and complexity of AI systems present novel challenges for trade secret protection, particularly when algorithms evolve beyond human understanding. It was highlighted that proprietary algorithms, training data, and chip configurations are all vital components requiring protection, especially from state-backed actors[ 00:32:45 ] .

China's Role and Competitive Practices

A significant portion of the discussion centered on China's aggressive strategies in the AI domain, with concerns raised about its intent to overtake the US through theft, deception, and systematic targeting of America's innovation base. China's rapid AI advancements, such as the DeepSeek model, were often attributed to intellectual property theft, reverse engineering, and exploiting legal gaps, rather than purely organic innovation. Specific tactics included cyber intrusions, insider recruitment, deceptive investment strategies, and weaponizing the legal system against American firms. While some acknowledged that DeepSeek might have used common distillation techniques, the consensus was that China engages in significant intellectual looting. Export controls on advanced AI chips were identified as a crucial, though imperfect, tool in slowing China's progress.

Balancing Transparency and Innovation

The hearing explored the delicate balance between mandating transparency for AI systems and protecting the trade secrets that fuel innovation[ 00:32:27 ] . Several speakers cautioned against "sweeping AI disclosure mandates" under the guise of transparency, arguing they could stifle innovation and inadvertently aid foreign adversaries[ 00:33:54 ]

. It was suggested that transparency should focus on specific areas like risk management, safety testing, and system capabilities rather than proprietary data or algorithms, which can often be shared with the government through secure channels or independent auditors under Non-Disclosure Agreements. The potential for well-designed regulation to actually support industry by building consumer trust and ensuring quality was also noted[ 01:16:06 ] .

Human Capital and Immigration's Impact on AI Leadership

A recurring theme was the critical role of human capital, particularly skilled immigrants, in maintaining US AI leadership. Concerns were raised that policies making it difficult for top international talent to stay in the US, such as threats to the H-1B visa program or visa uncertainties, could severely undermine American innovation. It was emphasized that a significant portion of AI startups were founded by immigrants and that foreign-born talent constitutes a large percentage of AI professionals at the PhD level[ 01:51:43 ]

. The ability to attract and retain the world's best and brightest was identified as a major asymmetric advantage for the United States in its competition with China[ 01:51:08 ] .

Role of Government and Regulation

There was a discussion about the appropriate extent of government involvement in AI development and security, with some advocating for treating AI companies as critical infrastructure to ensure minimum cybersecurity standards and information sharing. Recommendations included expanding security collaborations between the US government and AI companies, enhancing transparency through existing or new authorities, and investing in secure technologies and infrastructure. Suggestions for strengthening protections against foreign IP theft included codifying outbound investment screenings, increasing cyber requirements for federally funded tech firms, and requiring public disclosures for foreign ownership in American AI companies. The reauthorization of CISA 2015 was supported as a model for encouraging information sharing between the private sector and government.

Tone of the Meeting

The overall tone of the meeting was serious and urgent, reflecting a shared concern for US leadership in artificial intelligence amidst a global "AI arms race"[ 00:32:18 ] . While there was broad bipartisan agreement on the importance of protecting American intellectual property and innovation, particularly against threats from China, there were nuanced discussions and some disagreement on the precise methods and extent of government regulation and transparency mandates[ 00:42:09 ]

. The urgency was palpable, with warnings that the US has a limited timeframe to address critical gaps before risking its technological and strategic edge.

Participants

Transcript

Ms. Helen Toner

The subcommittee will come to order. Without objection, the chair is authorized to declare a recess at any time. We welcome today our guest for a hearing on trade secrets in the global AI arms race. I now recognize myself for a short opening statement. Today's hearing addresses the core tension in the age of artificial intelligence. the push for disclosure balanced with the need to protect innovation that drive US leadership. There's a non-disclosure on the phone. The proprietary algorithms that in fact train data sets are likely to be, now and in the future, the secret sauce. Not patented, not available to the public, but of such value that other nefarious actors will do anything to get them. Without strong protections, we risk sacrificing the very breakthroughs we aim to encourage. Some have called the sweeping AI disclosure mandates in the name of transparency. This committee has strongly believed and will continue to believe in transparency. But like anything else, a protective order is just a promise and a promise that is historically not kept, in fact, has no value. We cannot require disclosure of trade secrets or proprietary data sets under the guise of transparency. Such disclosures would not only have a chilling effect on innovation, but hand strategic advantage to foreign adversaries who do not play by the same rules. Consider this. The United States' investment in AI is approaching a trillion dollars. China has invested not more than one third as much and is already neck to neck with the United States.

Ms. Helen Toner

Deep Seek and other Chinese models are rapidly catching up to American companies. But let's look deeper. DeepSeq contains models that clearly were taken from OpenAI and other sets. We know China's speed is not merely because they invest in innovation. It is also the result of IP theft, reverse engineering, efforts to circumvent US expert controls, and more. China has also weaponized our legal system by financing lawsuits that target American firms in expensive litigation, and with the intent in many cases to mine that proprietary information. All the while, China continues to refuse meaningful protections on USIP in its own courts. I personally have been part of that recognition that they will not defend trademarks, patents, even when you go to the expense of getting them in China. Even if you make your product in China, you will not be protected. Lastly, American companies have a reverse situation, one in which we are working constantly, and this committee has worked tirelessly on a bipartisan basis, to promote rocket dockets, to in fact get the speed of decision to the lowest cost and the quickest time. Well, in China, you can be assured if you're an American company asserting your rights, you will go into the never never land of your case will come up someday. Efforts continue to include the need to prevent US engineers with access to specific sensitive AI systems from sidestepping non-competes and NDAs by working for Chinese firms. ensure the disclosure of regimes do not inadvertently reveal protection by mirroring eu style regulations and i might say here a little off the record eu as we speak is second guessing their very disclosures it is not it has not escaped us that the ai act of europe

Ms. Helen Toner

in fact, was a great step forward. And as we speak, many are pushing to take at least a small step backwards. And with that, I'm delighted to introduce my ranking member for his opening statement.

Darrell E. Issa

Thank you, Mr. Chairman. 117 days ago, Chinese company DeepSeek launched its flagship model, DeepSix R01. In the following days and weeks, American experts wrung their hands and shook their heads, many considering for the first time the possibility that we may be losing the so-called AI arms race. DeepSeek claims R01 was developed faster and cheaper than comparable U.S. models. Others argue DeepSeek, at a minimum, violated OpenAI's terms of service to obtain proprietary training data. Where everyone seems to agree is that AI startups should be innovating faster I've heard colleagues suggest we should forget regulations, ignore IP laws, and just focus on clearing the way for AI companies. I agree everyone wins when we foster American AI startups, but I think it's a false choice to say we can succeed only if we do so irresponsibly. Excuse me. If you walk around the Capitol today, you can see plants sprouting out of the soil beginning to grow. What you'll notice if you look closely is some of those shoots growing in the shade look like they're growing faster than the others.

Sign up for free to see the full transcript

Accounts help us prevent bots from abusing our site. Accounts are free and will allow you to access the full transcript.