AFP via Getty Images
The U.S. will step up its crackdown on China’s surveillance industry this coming week, with further restrictions on the supply of American technology that will be in effect from January 6. The new regulations target the use of artificial intelligence in geospatial applications—essentially the detection and classification of objects from planes, drones and satellites. While many such applications are civilian in nature—geographical surveys, construction, town planning, the real focus is military and surveillance. Essentially, the U.S. does not want these technologies in enemy hands. The restrictions cover U.S. exports to all countries bar Canada, but the primary target is of course China and its world-leading AI surveillance industry.
Some three months ago, Washington barred U.S. companies from trading with China’s leading AI surveillance unicorns—SenseTime, Megvii, Yitu and iFLYTEK, as well as its leading camera manufacturers, Hikvision and Dahua. That move was specifically designed to curb those companies providing technology and services to the authorities in China’s Xinjiang region, where an infamous surveillance state has been deployed to monitor the minority Uighur population. The Commerce Department cited “human rights violations and abuses,” including “mass arbitrary detention.” Again, though, those restrictions served to fetter China’s national-scale surveillance investments.
In reality, these new restrictions are much more specific—certainly more so than the industry feared: geospatial imagery software designed to train neural networks, with inherent espionage and military potential. Leading U.S. technology companies have been criticised before for AI collaboration with Chinese industry and research centres, what we’re seeing now is the start of a formal delineation as to what is off limits.
As reported by Reuters, this is the first set of regulations “to be finalized by the Commerce Department under a mandate from a 2018 law, which tasked the agency with writing rules to boost oversight of exports of sensitive technology to adversaries like China, for economic and security reasons.” The 2018 Export Control Reform Act formed part of the National Defense Authorization Act, under which a wide range of restrictions on China’s technology giants have been deployed, including prohibitions on the U.S. government purchase of equipment from Huawei.
No country spends more on advanced surveillance technologies or invests more in AI startups and growth businesses than China. One of the criticisms has been that the country has provided an unfettered surveillance laboratory for those domestic giants to hone their skills and their technologies without the usual constraints. The irony that America’s global technology players inadvertently feed into that industry with the complex sets of relationships between U.S. commercial players and China has prompted serious concerns from U.S. lawmakers.
Realistically, although this first export ruling is specific and limited in nature, you can expect to see further limitations on the supply of America’s AI technology to foreign states, with China in mind. While the “phase one” trade deal has taken some of the heat out of recent exchanges—as has the shift in focus to the Middle East, there remains a technology “Cold War” in place, with AI front and centre. There is also increasing pressure on the U.S. (and other Western countries) to clampdown harder on Chinese companies continuing to sell into Xinjiang. In fact, it would not be surprising if some of China’s technology giants, perhaps even Huawei, reviewed their approach to those sales in 2020, given increasing international pressure.
Disclosure: My company Digital Barriers develops AI-based analytics and surveillance technologies. This technology is not sold in China, but the company does compete from time to time with China’s surveillance companies in other countries.