Most people would like to be able to choose traditional methods of identification over facial recognition technology, the first major public opinion survey in China has shown, reflecting increasing concern in the world’s second-largest economy about personal data protection.
China has led the rollout of facial recognition globally, installing face scanners in transport hubs, schools, shopping centres and residential compounds.
But while 60-70 per cent of people believe the technology makes life safer and more convenient in those settings, users are concerned about their personal information being leaked and want more control over their data, according to results released on Thursday by the Nandu Personal Information Protection Research Centre in Beijing.
The survey highlights how the proliferation of facial recognition in China has created widespread concern and even resistance. The survey found that 74 per cent of respondents want the option to choose traditional ID methods over facial recognition. The survey covered online 6,152 respondents from October to November.
In October, China’s courts received their first challenge to the commercial use of face scans, and last week China’s education ministry was forced to respond to an outcry over the use of cameras in classrooms to track behaviour.
The number of people from Xinjiang whose data were exposed
According to the Nandu survey, the top concern was the possibility that operators of facial-recognition systems might be lax at data security and thus leak personal information, with 80 per cent of respondents identifying this among a list of concerns.
In addition, 57 per cent of respondents were concerned about their movements being tracked, while 84 per cent of respondents wanted to have the opportunity to review the facial-recognition data collected from them, or request that they be deleted.
The speed of China’s facial-recognition rollout has led to sometimes sloppy security around the online databases in which personal data are stored, security researchers have found.
Earlier this year, the location data of 2.6m people in Xinjiang were exposed in an unsecured database belonging to SenseNets, a facial-recognition technology company that contracts with police agencies.
The survey also highlighted the ways in which facial recognition systems often malfunctioned, and consent was not adequately gathered from users.
Of those who had encountered facial recognition, about 60 per cent had met with cases where it could not accurately identify them. Roughly a third said they had not signed user agreements or privacy policies before using face scanners, including in schools and residential areas.
The results come a week after SenseTime, one of China’s leading facial recognition companies, admitted that the public were paying increasing attention to the industry’s privacy and security problems.
SenseTime is now leading a consortium of 27 companies to write national standards on the use of facial recognition, under the guidance of a government body. The company is also on a US sanctions blacklist for its alleged role in aiding the oppression of Muslims in Xinjiang.
While domestic concerns grow, China’s facial recognition companies are leading the global market for public surveillance systems. They are also attempting to set international standards through the UN’s International Telecommunication Union, which often influences the technologies adopted by developing nations.