Megvii was the first facial recognition company in China to issue guidelines on the technology’s technical security and ethics in July last year. The move followed a similar initiative from Google’s Pichai, who published a list of AI principles in the wake of the now-abandoned Maven project. 

Megvii later promised in its IPO prospectus to establish an AI ethics committee. The company declined to comment further on its role in facial recognition regulation.

Other companies have joined the chorus. SenseTime, another major Chinese AI company, took the lead of a national standardization group for facial recognition technology last fall. The group includes other industry heavyweights like Alibaba affiliate Ant Financial, Xiaomi, Dahua and iFlytek. And state-backed AI company CloudWalk is participating in an AI criteria project led by the Chinese government, according to Shang.

However, company leaders and legal experts note that regulating facial recognition will be tough if the government doesn’t clear up legal uncertainties. The lack of any laws governing such technology in China allows companies to gain access to vast quantities of personal data.

The government has made some strides in this space with legislators drafting a new law to safeguard data privacy. But it has also sought to keep discussions about facial recognition limited to protecting user data and preventing things like identity theft and deep fakes, which popped into public consciousness after face-swapping app Zao went viral.

These are pressing problems for users. As the country has rushed to stick facial recognition into subway systems and toilets, China’s poor protection of biometric data has led to multiple data leaks. In one recent case, a middle school left facial recognition data of students and staff exposed in an unsecured database.

But much less has been said about the government’s use of facial recognition in surveillance and the role AI companies play in it. China has some of the most surveilled cities in the world, according to research from Comparitech. And it’s a big problem, according to Liu Deliang, a law professor at Beijing Normal University.

“Because it’s about regulating the government’s abuse of power, so it’s not only a law issue but a constitutional issue,” Liu told the South China Morning Post.

The issue is also becoming a problem for the country’s tech companies. Unlike Microsoft’s Smith, who has called for regulation against surveillance and discrimination, and Google’s decision to hold off on facial recognition projects, Chinese tech companies have been careful to avoid such topics. But this hasn’t saved them from scrutiny.

In October, the US put 28 Chinese public security bureaus and companies on a US trade blacklist over Beijing’s treatment of Uygurs and other minorities in China’s northwestern Xinjiang province. The list includes some of the biggest names in AI, including Megvii. Other listed companies include Yitu, another facial recognition company that’s reportedly considering an IPO, and iFlyTek, the national champion of voice recognition tech. The companies said they oppose the ban.