The Chinese version of TikTok, the - viral video app with more than two - billion downloads, is using facial recognition to censor live streaming by foreigners and children, raising concerns about potential use of the technology in the West.
Documents from parent company ByteDance describe how the Douyin app checks users' faces match their state ID cards before letting them stream.
It is part of a system that excludes foreigners and people from Hong Kong, as Beijing's digital censorship apparatus grapples with social media.
The technology continuously scans users' faces during their live streams to check that they are above the age of 16, while analysing their use of the app for signs of juvenile interests, such as cartoons.
Reports indicate that Douyin may be using the same technology to spot -foreigners who show their faces in Chinese users' broadcasts, cutting them off midstream if they refuse to get out of shot after a warning.
The system is likely to raise questions about whether TikTok, which is separate from Douyin but identical in many ways, might be applying the same technology in countries such as the UK and US.
TikTok is already under investigation by US regulators for allegedly violating children's privacy, and is being sued in the state of Illinois by four teenagers who claim it collected their facial biometric data without asking for their consent. Foreigners in China have officially been banned from streaming without a government permit since 2017, but that rule appears not to have been enforced until recently.
A spokeswoman for ByteDance confirmed that it had censored Joshua Dummer, a British man living in Beijing, who had briefly appeared in a live stream by his wife, who is Chinese. The company said: "As all live streaming moderation on Douyin is conducted by human moderators in real time, we can confirm that the action in this case was carried out by a person.
"It declined to say whether facial recognition was involved, but its own documents show it routinely uses various artificial intelligence techniques to flag suspicious streams to its human moderators. Those systems are just one component of a "complete monitoring process" developed in response to a live-streaming boom that has minted new social media millionaires but strained China's censorship apparatus.
A spokesman for TikTok declined to say how the app used facial recognition, noting only that it used human moderators to "evaluate" live streamers' age. He said: "TikTok takes the safety of our younger users seriously... TikTok has never provided user data to the Chinese government, nor would we if asked to do so.