CNAS Responds: Outbound Investment Executive Order

Today, the White House released an executive order (EO) detailing a ban on U.S. private-equity and venture-capital investments in some Chinese technology companies with potential military applications through a series of outbound investment mechanisms. Experts from CNAS explain what is included in the EO, and what it means for national security and U.S. companies.

All quotes may be used with attribution. To arrange an interview, email Alexa Whaley at [email protected].

--

Emily Kilcrease, Senior Fellow and Director, Energy, Economics, and Security Program:

The President has signed the long-awaited EO on outbound investment, accompanied by a notice of proposed rulemaking from Treasury. There are few surprises for those following this debate, including the tight scoping of covered technologies and a focus on “smart money” transactions. Today’s announcements did confirm a few important details, notably that there will be strict prohibitions on investments related to chips, quantum, and – potentially – AI. AI is the area where the administration seems the most uncertain, as the EO lacks a clear vision for what types of AI are most important for U.S. national security interests. The proposed coverage of AI transactions likely leaves a lot of the most exciting – and potentially risky – advances out of scope.

This EO will make many camps unhappy. China hawks, both in Congress and former Trump administration officials, wanted more holistic decoupling of investment ties. Industry will chafe at the additional roadblocks for investing in an increasingly difficult market. Traditional economists will argue that China has plenty of access to capital from sources other than the United States (though this risks confusing macro data with micro problems).

The administration is trying to throw a pitch straight down the middle, and with this EO they have largely succeeded. The scope hits on those transactions that present high national security risk, without burdening agencies with a program that is too big to properly enforce. It’s a targeted approach that has a strong grounding in national security concerns, rather coming off as protectionist. Importantly, it should form a strong basis for further dialogue with partners and allies on standing up similar mechanisms, which will be critical to the broader economic security agenda.

Keep an eye on Chinese reactions. Chinese complaints would be highly hypocritical, given China’s longstanding and extensive restrictions on both inbound and outbound investment. But, the release of the EO could complicate recently revived bilateral talks regardless. If China maintains the current tempo of talks while refraining from retaliation, it will be a sign that both countries are interested in responsible management of the world’s most consequential economic relationship.

Travis Mosier, Adjunct Senior Fellow, Energy, Economics, and Security Program:

In the context of China’s semiconductor industry development, the new executive order (EO) requiring notification or outright restricting some U.S. investment will make it even more difficult for the PRC to realize its ambitions of semiconductor self-sufficiency with its self-evident national and economic security benefits. Even under ideal conditions including ample access to capital, barrier-free trade flows, and efficient supply chains, the semiconductor industry is complex and highly competitive. While the U.S. rightly and narrowly tailored the restrictions to advanced semiconductor and microelectronics technology, it represents an additional hurdle to China developing globally competitive semiconductor technologies and products.

More broadly, the new China investment restrictions are yet another symptom of the deteriorating US-China relationship, and no matter how “small” the yard is, the EO will be viewed by the Chinese Communist Party (CCP) as further evidence of the U.S. strategy to contain China’s rise, especially its ambitions in the semiconductor space. As witnessed in the wake of the Oct. 7 export controls, any retaliation will be limited as China continues to walk the razor’s edge between welcoming foreign investment while addressing and downplaying the realities of its struggling economy, growing instability within the CCP, and its deteriorating geopolitical environment and reputation. One area that the US could see retaliation is an expansion beyond recent gallium and germanium export restrictions to include other critical inputs of the semiconductor manufacturing process.

Thomas Krueger, Adjunct Senior Fellow, Energy, Economics, and Security Program:

The Outbound Investment E.O. is intended to complement the Administration’s October 7th semiconductor export control regulations that targeted advanced semiconductor manufacturing, PRC Artificial Intelligence, PRC supercomputers and Quantum Computing that could enable Xi’s military capabilities. Like, the October 7th rule, the E.O. is targeted to restrict investments in advanced technology. Where the export control measures targeted export, re-export and transfer of the items themselves, and to some extent, activities of U.S. persons with respect to advanced semiconductor manufacturing; this E.O. sets the stage to fill that investment gap that Commerce authorities do not cover.

There are a few notable differences in the scope of the technologies captured between the Outbound Investment E.O. and the October 7^th rule. First, the national security rationale is articulated differently than in the Commerce rule. In the October 7^thrule, the national security rationale focused on how the technologies listed support the PRC’s Weapons of Mass Destruction systems and surveillance systems that can enable human rights concerns. The clear explanations Advanced Notice of Public Rule Making provide meaningful clarifications of the breadth of the US government’s concerns with different technologies. For example, the U.S. is concerned with AI driven penetration testing software that support the military, intelligence, or mass-surveillance. In a stark difference between export controls and this E.O., outside of the entity list, there is not a general prohibition on penetration testing tools to the PRC (AI-enabled or not). Also, there is no mention of Human Rights directly, though "mass-surveillance” in a sense replace “human-rights”. This may be a tip to partners and allies that are not yet comfortable using “human rights” as as discriminator to restrict investments. Also, quantum is mentioned as a prohibition. This is not surprising considering that quantum was called out as a “force multiplying technologies” in September of 2022 by National Security Advisor Jake Sullivan. However, at this time, there is a no meaningful export control measures on quantum; and it is uncertain how far the Administration will go to leverage export controls on quantum considering how nascent quantum is as a technology.

John Hughes, Adjunct Senior Fellow, Energy, Economics, and Security Program:

The outbound investment EO expands the coercive economic measures toolkit available to the Biden administration, adding a potentially potent new tool on top of sanctions, export controls and CFIUS reviews to help advance foreign policy priorities.

The EO is narrowly tailored to focus on certain key technology areas where withholding US investment could impact the ability of countries of concern to develop capabilities that threaten US national security. This approach reflects a balancing act between wanting to have impact while not unnecessarily limiting broader investments. At the same time, it is almost certainly not the last word and will likely evolve over time as is evidenced by the lengthy comment period and detailed questions contained in the ANPRM, as well as the specific requirements within the EO for the Secretary of the Treasury to assess the effectiveness of the accompanying regulations and determine whether to shift the scope within one year of their effective date.

We can expect to see changes to the scope and focus as the administration receives data and feedback on its impacts, with current areas of focus expanded, tweaked, or even potentially abandoned to match priorities and concerns. This will be an important process to make sure they get it right - unlike the other tools in the toolbox, this is brand new territory with potentially unforeseen and unanticipated consequences. Getting it right will require flexibility and a willingness to thoughtfully consider the input from the investor community and others that are directly impacted.

Jacob Stokes, Senior Fellow, Indo-Pacific Security Program:

The executive order (EO) placing controls on some outbound U.S. investments in China while requiring notifications for others marks a new stage in the Biden administration’s prosecution of strategic competition with Beijing. The EO further demonstrates America’s determination to ensure that U.S. capital does not support China’s development of critical and emerging technologies to modernize its military, the People’s Liberation Army; enable human rights abuses; and otherwise advance China’s geopolitical ambitions. The underlying rationale is clear and justified: The two major powers are engaged in a geopolitical contest spanning the security, technology, and governance domains. At a minimum, U.S. entities should not fund (or advise) R&D or commercial activities that could help China gain a technological advantage in those arenas.

China will surely denounce the EO as just the latest American action designed to hamper Beijing’s overall technological development, part what Xi Jinping has called “comprehensive containment, encirclement, and suppression” by the United States. Beijing will claim the EO runs counter to recent bilateral diplomacy aimed at stabilizing the U.S.-China relationship. The Biden administration should—and likely will—listen to but ultimately reject China’s arguments. The EO is a smart step that will enhance the United States’ ability to compete effectively with China in the years to come. Having an outbound investment control regime in place—and perhaps even strengthening it over time—will only become more vital as the technologies it covers, especially artificial intelligence and quantum computing, mature. Moreover, for U.S.-China diplomacy to be successful going forward, the pair will have to manage ties amid competitive actions taken by both sides rather than proceed only when all disputes have been settled.

Tim Fist, Fellow, Technology and National Security Program:

What can we say about the probable effects of restricting U.S. private investments in the targeted technologies (semiconductors, AI, quantum)? And will the proposed restrictions be useful for AI in particular?

On the probable effects of investment restrictions, one relevant observation is that these technologies are IP-heavy. For advanced semiconductors, the U.S. (and key allies such as the Netherlands and Japan) have quasi-monopolies over key chokepoints in supply chains because only a small number of firms have the technical know-how to build the incredibly complex tools required for the industry, such as extreme ultraviolet lithography machines. For AI, the current frontier of R&D is dominated by a handful of U.S. labs (and their big-tech backers) with the technical know-how to train gigantic models using vast quantities of data and computing hardware, and to industrialize modern AI techniques like reinforcement learning from human feedback (RLHF). Quantum computing is an earlier-stage technology without a clearly dominant technical paradigm, but IP such as ion traps, dilution refrigerators, and error correction software for quantum algorithms could be important. Given that the PRC and Chinese private investors will likely be willing to spend big on these technologies where they show promise (for example, a $143 billion package is reportedly in the works for China's semiconductor industry — compare this to the CHIPS and Science Act at $53 billion), investment restrictions seem most useful for U.S. national security interests in cases where they prevent IP transfer from key industry players to China, rather than the direct effects of the investments themselves. This is likely a solid bet to make: private equity investments are often paired with expertise to help the target firm succeed in their industry. If we look at the draft regulations, this is indeed the bet that Treasury appears to be making: they explicitly call out categories of sensitive IP in each of the three tech areas being targeted.

Within the AI domain, Treasury currently looks unsure about whether and where they’ll proceed with restrictions. The proposed regulations state that “If the Treasury Department were to pursue a prohibition in this category, a potential approach is to focus on U.S. investments into covered foreign persons engaged in the development of software that incorporates an AI system and is designed to be exclusively used for military, government intelligence, or mass-surveillance end uses.” They also consider a lighter-touch notification-based regime for investments into AI systems used in cybersecurity, robotics, and smaller-scale surveillance. This language excludes the highly-capable general-purpose AI systems that represent the current frontier of AI R&D. This will likely be a focal point of public comments over the next month and a half: while today’s frontier models are generally commercial in nature, they also show significant dual-use potential in areas such as cyber offensive operations and biological/chemical weapons design.

Hannah Kelley, Research Associate, Technology and National Security Program:

The new executive order doubles down on what Washington has been saying for some time: technological leadership is a critical indicator of global power, key inputs including talent, materials, equipment, and capital investment enable technological ascendency, and so the United States should be mindful of how its own inputs are bleeding into untrusted environments and/or hands—especially in countries of concern like China.

When we think about high-tech sector U.S. capital investments in adversarial ecosystems, the enabled development of otherwise benign capabilities can still increase baseline critical infrastructure, technical know-how, and the ability to more rapidly scale nefarious capabilities on the side or in the future. So while concern over long-term capacity building may not require immediately cutting off U.S. investment across the board, it is ample reason to monitor developments in these spaces. Meanwhile, instances of more clear U.S. capital diversion that in turn threatens U.S. national security—for military advantage, oppression and surveillance networks, etc.—demand immediate action.

I’m interested to see how the Administration negotiates investments to monitor vs. investments to squash when it comes to the discrete capabilities and supply chain dynamics of these three broad sectors. I’m likewise curious if/when biotech might get added to this list. But on the whole, I think investigating the modern impacts of outbound investments is long overdue if the United States is serious about the relevance of tech for global power and the growing relationship between economic and national security.

Sam Howell, Research Assistant, Technology and National Security Program:

The Biden administration’s decision to include quantum technologies in its outbound investment EO underscores the potential strategic significance of these capabilities. Quantum computing, for example, promises to catalyze transformational breakthroughs in a range of industries, from medicine and financial services to clean energy and defense. But quantum computers are still relatively immature, and it could be years or even decades before researchers and engineers overcome the technical hurdles currently preventing quantum computers from delivering real-world effects. Introducing outbound investment regulations on the quantum industry now, at such an early stage of technology development, sends a clear message: the United States cannot afford to fall behind competitors in the global race to build and scale functional quantum systems.

The outbound investment mechanisms laid out in the EO could help U.S. policymakers track the exchange of quantum technology components and expertise and gain greater visibility into competitors’ quantum activities. At the same time, the EO could unintentionally stifle the nascent U.S. quantum industry just as it is beginning to grow. Close public-private cooperation and communication during the implementation phase will be critical to ensure that the EO achieves its intended outcomes without imposing undue burdens on the U.S. quantum ecosystem.

--

All CNAS experts are available for interviews. To arrange one, contact Alexa Whaley at [email protected].