Sanctions Scrum November 2022: Crippling Sanctions – US moves decisively to hobble Chinese AI ambitions. Move reflects new policy with broad implications.
Dear Colleagues and Friends,
When mud and blood mix in your mouth it tastes like tinny tapioca pudding.
I know this thanks to getting knocked out on a pitch in Gloucestershire, England. One moment I was kneeling on the sideline—having just been tackled, I had been quietly re-inflating my lungs with oxygen. The next moment, I was staring up at the tower of the Tewkesbury Abbey. My head hurt, my lungs were empty again, and my mouth tasted like tinny tapioca.
Imagine my surprise to learn that I had been attacked by my own teammate! I sensibly presumed I was targeted by an opposing rugger in our home “friendly” match with Cheltenham; but, oh no, I had been kicked-cold by my own mate, Nigel! In his modest mind, Nigel was teaching “the Yankee” an important lesson: “when tackled, don’t linger; if you rest on a knee, someone will kick you in the head!”
Don’t dawdle in the dirt. Pop-up, young rugger! Pop Up!
The Biden Administration’s unprecedented assault on the Chinese AI sector is being described in morbid terms. The U.S. is variously being accused of “choking”, “kneecapping”, and “strangling with an intent to kill” the Chinese market for AI-enabling microchips. The consensus metaphor is “crippling”.
When a fifteen-stone rugger throws his knee into your head at speed, in my experience, you are crippled. But what does it mean to suffer crippling sanctions? And what lessons can we learn from the comprehensive U.S. effort to hobble Chinese AI ambitions?
Whether you’re a member of the board of directors, a corporate officer—or just a curious executive trying to understand where to stick your bandaged head—we’re here to help.
Thanks for joining us on the pitch! Now let’s bind up and win this thing!
Best,
Jim Lay
Philomont, Virginia
26 November 2022
________________________________________________
Crippling Sanctions: US moves decisively to hobble Chinese AI ambitions. Move reflects new policy with broad implications.
US REGULATORS recently moved to hobble the Chinese AI industry. [1] Global business leaders might dismiss the moves as isolated. That would be a mistake. [2]
Apoplectic anthropomorphisms abound. Wired magazine reports the US has “kneecapped” [3] the Chinese chip sector. Normally staid think tanks report of the U.S. “choking” China’s access to the future of Artificial Intelligence (AI) [4] Unsated by all this carnage, U.S. legislators seek more “crippling” sanctions still. [5]
The hyperbole may be justified. The new rules seek “to keep [China’s] AI industry stuck in the dark ages while the US and other Western countries advance.” [6] Whatever your metaphor of choice, the intent of the new rules is to unapologetically hobble a critical Chinese technology sector.
As a political idea, crippling sanctions sound good. After all, anything short of kinetic war means lives saved. So, a little crippling should be encouraged, provided its effects are limited to our adversaries’ technology supply chain, right?
But it turns out that crippling a large, developed economy is hard. Whatever the political benefits of deploying the white-hot rhetoric, the substantive case for embargo sanctions is weak. “The biggest flaw,” The Economist observes, “is that full or partial embargoes are not being enforced by over 100 countries with 40% of world GDP.” [7]
This doesn’t bode well for containing China’s technological ambitions. The lesson is that marginal actors will sit-out embargo sanctions, diminishing their effectiveness. [8]
In rugby, if you can’t field a complete team, you forfeit the match and head to the pub. In the collective mind of the Biden Administration, if you can’t get everyone heaving together, you go it alone.[9]
When compared to a multilateral embargo, unilateral, sector-specific sanctions are a breeze. It’s simpler to cripple the specific supply-chain ecosystem causing you the most grief. And a narrow aperture limits unintended damage to unrelated industries, entities, and persons. In theory, at least, sector-specific sanctions minimize collateral damage; the average Chinese citizen is unaffected by U.S. restrictions on 7-nanometer microchips.
The new rules will challenge this orthodoxy. The scope is unprecedented in modern sanctions practice. [10] By design, the new rules seek to decimate narrowly and deeply. But, once deployed, new sanctions tools seldom gather dust. And they seldom remain limited in application. Global innovators are wise to pressure test [11] their controls; in the days ahead, the U.S. is likely to find additional strategic technologies quietly kneeling on the side of the pitch. [12]
Greg Allen is a Senior Fellow at the informed Center for Strategic and International Studies (CSIS). He sees it this way:
There are four interlocking elements of the new policy targeting different segments of the semiconductor value chain, and all elements must be understood simultaneously to grasp the scope of what the Biden administration plans on achieving. In short, the Biden administration is trying to (1) strangle the Chinese AI industry by choking off access to high-end AI chips; (2) block China from designing AI chips domestically by choking off China’s access to U.S.-made chip design software; (3) block China from manufacturing advanced chips by choking off access to U.S.-built semiconductor manufacturing equipment; and (4) block China from domestically producing semiconductor manufacturing equipment by choking off access to U.S.-built components. [13]
According to CSIS, “These actions demonstrate an unprecedented degree of U.S. government intervention to not only preserve chokepoint control but also begin a new U.S. policy of actively strangling large segments of the Chinese technology industry—strangling with an intent to kill.” [14]
Choking? Kneecapping? Strangling with an intent to kill?
What’s a Global Rugger to do?
Tackling Tips
Global innovators able to insulate themselves from risks associated with Chinese technology dependencies should do so. That’s the smart, long-term play for sophisticated Western actors and their global sympathizers.
Those unable to insulate their supply or service chains should take a page from Deputy Attorney General Lisa Monaco’s playbook. Readers will recall that Monaco admonished us to “pressure test” our systems and controls. [15]
To do this, imagine what it looks like in your corner of the global economy if the U.S. government put a target on the back of your supply chain dependency. What if your value chain was the next focus of the new U.S. policy to “preserve chokepoint control”? [16]
To design your test, look to the new rules and how they seek to degrade Chinese AI. The new rules seek to do each of the following: (i) deny access to the technology at issue; (ii) deny access to software that could allow the target nation to design the technology; (iii) deny the target nation from acquiring machinery that could manufacture the technology domestically; and (iv) deny access to any relevant US made elements of the technology.
The pressure test needs to be able to scale to treat dynamic targets. One of the early lessons learned from the Huawei export controls is that the initial effects quickly leach to numerous target affiliates. This prompts the target to spin-up additional affiliates to obfuscate the source of products or services. And this creates a cycle of new affiliates begetting new targets. Put simply, the individuals and entities subject to the controls is large and fluid. Picture squeezing Jell-O in your fist.
Our testing therefore needs to (i) prepare for a rule which addresses multiple segments of the implicated technology value chain, and it needs to (ii) scale to account for a dynamic scope, ever evolving to identify novel entities used by the sanctions target to evade enforcement.
These design goals inform three practical tips:
First, simplify. The foundation of any robust control is a Sanctions Control Program (SCP) mimicking the Framework for OFAC Compliance Commitments [17] and thoughtfully tailored to your own, idiosyncratic risk. OFAC “strongly encourages” an SCP for “organizations subject to U.S. jurisdiction, as well as foreign entities that conduct business in or with the United States, U.S. persons, or using U.S. -origin goods or services.” [18] Confirm your SCP is properly designed, successfully deployed, and adequately resourced to identify, escalate, adjudicate, and report suspect transactions.
If you sit on a Board, you have a fiduciary duty to bake management’s commitment to its SCP into your company’s DNA. For officers, failing to promote an SCP is professional malpractice.
Second, leverage technology. Modern sanctions practice requires the efficient adjudication of large sales-platform data against fluid, global lists of persons, entities, sectors, regions, and countries targeted for sanctions. Developed-economy sanctions coordinate imperfectly. Ambiguities between Allied controls, for example, can introduce risk.
U.S. Government guidance is informative:
Given the dynamic nature of U.S. economic and trade sanctions, a successful and effective SCP should be capable of adjusting rapidly to changes published by OFAC. These include the following: (i) updates to OFAC’s List of Specially Designated Nationals and Blocked Persons (the “SDN List”), the Sectoral Sanctions Identification List (“SSI List”), and other sanctions-
related lists; (ii) new, amended, or updated sanctions programs or prohibitions imposed on targeted foreign countries, governments, regions, or persons, through the enactment of new legislation, the issuance of new Executive orders, regulations, or published OFAC guidance or other OFAC actions; and (iii) the issuance of general licenses. [19]
As a practical matter, this Ask [20] requires a thoughtful integration of modern, third-party technology solutions.
Last, with critical components, develop redundant suppliers. Do so in each different segment of the implicated value chain. Here, too, the Huawei controls are instructive. Those uniquely exposed to them—primarily telecommunications providers, the IoT industry, government contractors, and other highly regulated actors—would report that even these early government efforts to hobble the Huawei ecosystem were expansive.
For example, the provisions of the Federal Acquisition Regulations (FAR) that codified the Huawei rules prohibited both (i) use of the targeted technology in government programs, and (ii) use by market participants generally and throughout their enterprise, independent of performance on a government contract. [21] This second requirement—that conforming companies purge their broader enterprises of the nefarious technology, independent of a U.S. government nexus—was considered an overreach at the time.
When compared to the new controls, the Huawei controls appear tame. Under the new paradigm, the U.S. government would have (i) sanctioned the use of all Chinese telecommunications equipment; (ii) denied China access to software that could allow it to design telecommunications equipment; (iii) deny China from acquiring machinery that could manufacture the telecommunications equipment domestically; and (iv) deny China access to any relevant U.S. made elements of Chinese telecommunications equipment. And this multi-segment approach informs supply chain redundancy.
As CSIS’s Greg Allen observed, “There are four interlocking elements of the new policy targeting different segments of the [targeted] value chain, and all elements must be understood simultaneously to grasp the scope of what the Biden administration plans on achieving.” [22]
New U.S. efforts to “preserve chokehold control” of Chinese AI inaugurate a powerful new government tool. Its stated purpose is to degrade the development of Chinese AI technology, but its future application is unclear. To prepare, Global Ruggers should pressure test their existing SCP, fully leverage technology, and develop redundant suppliers for critical supply chain inputs.
