[The princess’s suitors] were told to cut a slice of moon, or change the ocean into wine. They were set to finding things that never were, and building things that could not be.
—James Thurber, “The Thirteen Clocks” (1950)
In mathematics, a null set is a category with nothing in it, an empty box. By instructing the executive branch, led by the Department of Commerce, to identify and establish new export controls on “emerging” and “foundational” technologies, Congress may have mandated the creation of something approaching a null set, which might explain why there has been so little movement on it in the last year and a half.
Where does the process stand now? And why has it taken much longer than expected to define these technologies and set corresponding controls? Senators Tom Cotton and Chuck Schumer in November wrote to Secretary of Commerce Wilbur Ross demanding answers to these and other questions. But the task might not be quite as simple as Congress imagined.
How the system works
Our Department of Defense seeks to give our military the most advanced, most reliable weapons and other equipment that we can provide. On the flipside, our export control system seeks to ensure that our adversaries do not have comparable weapons and equipment. Export controls, set by Commerce and through multilateral regimes, do this by restricting the transfer of sensitive technology to those who might employ it against our allies or us.
United States export controls have been in effect continuously since shortly before World War II began, as well as—for weapons—on numerous other occasions since the nation’s founding. The newest US export control law, the Export Control Reform Act (ECRA) covers so-called dual use items—those having recognized civilian as well as military applications—and low-level military items.
Among other things, ECRA requires the executive branch to identify, and the Commerce Department to control, exports of emerging and foundational technologies that “are essential to the national security of the United States” and aren’t already subject to export controls. Those implementing ECRA have to consider national security—of course—as well as whether the technology is already widely available elsewhere, whether a unilateral control would harm domestic research and innovation, the impact of controls on the defense industrial base, and the willingness of allies to impose similar restrictions.
For a host of reasons, I question whether the results of this exercise will be materially different from what the existing system would have generated.
“Emerging” and “foundational”: opposite conundrums
No one doubts that the United States should control emerging technologies that have national security implications, or that such controls should be imposed as early in their development as practicable. Indeed, that is what the executive branch has been doing for decades.
Until an emerging technology is being applied in fairly specific ways, though, it is difficult to write regulations that are sufficiently precise to be meaningful to regulators and exporters.
A vague law is no law at all.
–US Supreme Court, United States v. Davis (2019)
By way of example, the Commerce Department can’t issue a regulation that says, “Don’t send any artificial intelligence technology to China” unless that regulation sets out particular applications and technical parameters. This is because a general prohibition isn’t specific enough to inform exporters what can and cannot be sent to China or elsewhere, or to inform even Commerce’s highly-trained enforcement agents, let alone lay prosecutors, judges, or juries, when an exporter may have broken the law.
The US Supreme Court reminded us recently that “vague laws contravene the first essential of due process of law that statutes must give people of common intelligence fair notice of what the law demands of them.” Moreover, and to paraphrase a 2000 Supreme Court ruling about sanctions, export controls must be “drawn not only to bar what they prohibit but to allow what they permit.” In short, our Constitution requires that a regulation set out clearly and specifically the boundary between what is lawful and what is not. That, in turn, requires the kind of specificity that one sees in the several hundred existing entries on the Commerce Control List.
Beyond legal considerations, when we unilaterally control any technology too tightly, there’s a good chance that we will drive research and development, and ultimately production, offshore. This is so whether the technology is emerging or well on the way to being in common use, and we have seen very tight US export controls play a role in engendering robust foreign competition in sectors like machine tools, commercial use of space, and commercial thermal imaging.
In this vein, ECRA correctly recognizes that for controls to be truly effective, they should also be adopted by our allies in the four multilateral export control regimes. Securing agreement for multilateral control is difficult, time-consuming work, but as history and ECRA indicate, it is the most promising route to success.
So, is ECRA anything new?
When it comes to controlling emerging technologies, the government has little choice but to do what it already has been doing for decades and what ECRA, read closely, is telling it to do now: follow emerging technologies closely, with a particular eye toward applications that could give a military or intelligence advantage to an adversary. When those potential applications begin to become concrete (and hence to be suitable subjects for legally enforceable regulations), control those—and, if at all possible, do so in the context of the multilateral export control groups rather than unilaterally.
Luckily, the one set of emerging technology controls published by Commerce is multilateral and specific, and sources indicate that future rules will be similar. But this takes time.
In a sense, foundational technologies sit at the opposite end of the spectrum from emerging technologies. While it can be too soon to control an emerging technology if specifics are not available, by contrast the problem with foundational technologies is that it may be too late to control them effectively. By definition, their uses are widespread—so much so that they are specific, well known, and typically available from numerous sources outside the United States. In many instances, most or all export restrictions on such technologies—unilateral as well as multilateral—have been removed or sharply curtailed.
A frequently cited example is that of semiconductor exports to China. China goes to great lengths to procure cutting-edge semiconductors and use them for military purposes. But those high-end items already are subject to tight, multilateral export controls, and China cannot obtain them legally.
But China also wants to buy large volumes of microprocessors and other commodities whose technology is several generations old, for use in consumer products in furtherance of its Made in China 2025 effort. These items, and the technology needed for their production, no longer are viewed as having significant military utility. For that reason, they may be subject to reduced controls, or even de facto decontrol, by the multilateral groups to which the United States belongs and by the United States itself.
This appears to be an issue of economic security, though ECRA implicitly disclaims this intention, mentioning only national security and foreign policy as the basis for new controls. To the extent that the relevant technologies are of US origin, the United States could try to re-control them and prevent the sale of the resulting commodities to China. It’s far from certain, however, that our allies would agree to do the same.
China generally prefers to purchase products that use US technology because our goods usually are the most reliable. If US-based supplies were to become unavailable, though, China would shift its purchases to other sources. Indeed, news reports indicate that Huawei, which was placed on the Commerce Department’s Entity List several months ago, already has begun finding new suppliers outside the United States.
The problem with controlling foundational technologies, then, is their ubiquity.
The problem with controlling foundational technologies, then, is their ubiquity. Simply put, the United States ordinarily isn’t the only potential source. Given that, preventing China from acquiring such items made here or based on our technology could end up hurting US companies, US workers, and our overall defense industrial base more than it impairs the Chinese effort to dominate us economically.
The Commerce Department is expected to publish an advanced notice of proposed rulemaking for foundational technologies in the near future. That may indicate the options that the executive branch is considering.
Input from industry
We saw in the course of Obama administration’s successful Export Control Reform (ECR) initiative that it is important to seek private sector input on proposed controls. Although the Commerce Department may set the rules, US tech firms are the ones on the frontlines, particularly with emerging technologies. With the most in-depth knowledge of the technologies—that they themselves developed—industry’s feedback is critical to developing effective controls on the right technologies.
In addition to having the technical wits, private firms are intimately familiar with the impact that export controls have on their competitiveness. Although government-employed technical experts are knowledgeable, they do not always have full information on what is currently available in the global marketplace, nor are they always able to predict the full range of consequences that new controls may have on US businesses. In the case of the ECR rules, input from industry helped ensure that the controls neither over-controlled nor under-controlled the technologies in question.
* * *
I have grave doubt whether the assignment to control emerging and foundational technologies will result in controls significantly different from what the existing system—which operates fairly well—would have produced. That said, it’s important to recognize the many positive aspects of the ECRA statute beyond its provisions on emerging and foundational technologies. Among other improvements, it codifies the changes made by the ECR initiative. It continues the long-standing tradition of using export controls to prevent the proliferation of weapons of mass destruction, strengthen our defense industrial base, and focus controls on items that pose a serious national security threat.
Make no mistake. The intentions of Congress in drafting ECRA’s provisions on emerging and foundational technologies were good, and Senators Schumer and Cotton do no wrong in pressing the point. The fact remains, however, that there likely will turn out to be less here than meets the eye.
Eric L. Hirschhorn served as Under Secretary of Commerce for Industry and Security from 2010 until 2017. In that capacity, he headed the Bureau of Industry and Security and was a leader of the Obama administration’s Export Control Reform initiative. Prior to that service, he spent three decades conducting a private law practice that included economic sanctions and export control matters. He is a retired partner of Winston & Strawn LLP.