iFlytek automated speech recognition system raises human rights concerns

A partnership between Chinese technology company iFlytek and China's Ministry of Public Security to develop an AI-powered system that automatically identifies targeted voices met with consternation from civil society organisations.

According to Human Rights Watch (HRW), the two parties intended to develop a pilot surveillance system that builds on the Chinese government's existing Automatic Speaker Recognition system to automatically identify targeted voices in phone conversations. 

Chinese media reports suggested the system will be applied for counterterrorism and 'stability maintenance' purposes.

Chinese police are thought to have collected approximately 70,000 voice samples by 2015. By contrast, the country's facial image database contained data on over a billion individuals. iFlytek made 80 percent of China's speech recognition technology, HRW said.

The finding raised concerns about the surveillance and privacy of Chinese citizens and ethnic minorities, notably Uyghurs and Tibetans.