US considers AI export controls in fight to keep lead against China

Proposed export laws would cover everything from AI to WMDs

The US Department of Commerce’s Bureau of Industry and Security recently proposed doubling down on export controls for conventional weapons, intelligence collection, weapons of mass destruction and items that could be deemed to have potential terrorist applications.  The move comes under the umbrella of the Export Control Reform Act established earlier this year, which aims to protect emerging and foundational technologies that are deemed essential to US national security; and comes on the heels of crackdowns on foreign investment in US emerging technologies.  Members of the public have been invited to share their opinions on the proposed export controls.

In framing the new export controls for a given technology, the Department of Commerce considers “the development of emerging and foundational technologies in foreign countries; the effect export controls may have on the development of such technologies in the United States; and the effectiveness of export controls on limiting the proliferation of emerging and foundational technologies in foreign countries.”  Along with the clear threat posed by conventional weapons and dual-use technologies with potential terrorist applications, the Trump administration is concerned about other countries capitalizing on US technology advances in order to spur their own technological research and development.

AI technologies considered a major asset to US national security

Prominent on the list of emerging and foundational technologies that may be considered important to US national security is the whole gamut of artificial intelligence and machine learning technologies, including neural networks and deep learning; computer vision; speech recognition and production; machine translation; deep fake and cloud technologies, among others.  The rules would affect leading tech companies like Google, Amazon, IBM and Apple.  It’s evident that this is a move designed to halt China’s advance in AI research and development, as it poses the biggest threat to the US’ current dominant position.

In light of the current trade war between the US and China, and China’s stated aim of becoming a leading manufacturing power with its “Made in China 2025” policy, the Bureau of Industry and Security is particularly interested in the development status of these technologies in the US as compared with other countries and the potential impact of the controls on US technological leadership.  The maturity level of the emerging technologies will be a factor in creating the guidelines for export control.  Trade talks between China and the US are expected to come to a head at the G20 summit in Argentina next week.

US will have a tough time controlling AI intellectual property

Of course, as we’ve seen in cases such as the deep fake algorithms that went viral last year, it’s not always so easy to halt the spread of new technology.  Despite the Trump administration’s efforts to curb China’s intellectual property theft such as Fujian Jinhua’s alleged appropriation of Micron’s trade secrets this month, open-source code will bleed across borders, there’s no way around it – so is the US just shooting itself in the foot if it decides to limit exports by its very own companies?

Where the US leaves a gap in the Chinese market, other countries with advanced technologies may fill it, leaving the US at a disadvantage.  While the US has traditionally led the AI scene in terms of research and development, China is catching up quickly thanks to the reams of data in its possession, which allows for faster machine learning.  Meanwhile, the US lacks a clear AI strategy and risks falling behind if it doesn’t organize itself soon.

Is the US right to implement export controls on AI and emerging technology, or would it do better to focus on its own policies surrounding AI?