Alibaba (BABA) says that it has stopped trying to identify faces by ethnicity after the company was accused this week of creating a facial recognition system meant to detect Uyghurs.
The Chinese tech company released its statement after The New York Times and IPVM, a surveillance industry publication, reported that a division of Alibaba showed its clients how they could use its technology to identify members of the Uyghur community in videos and images.
The system was built by Alibaba’s cloud computing team, and allegedly included the example: “Is this a Uyghur?” in an algorithm, according to the report. It also reportedly included code to recognize if someone was a “minority” or “Asian.”
In a statement Thursday, Alibaba said that it was “dismayed to learn that Alibaba Cloud developed a facial recognition technology in a testing environment that included ethnicity as an algorithm attribute for tagging video imagery.”
Alibaba did not mention Uyghurs in its statement, or explain how or why the system was built in the first place. But it stressed that the technology had been limited to trials, and “was not deployed by any customer.”
“We never intended our technology to be used for and will not permit it to be used for targeting specific ethnic groups,” the company said. “We have eliminated any ethnic tag in our product offering.”
China has long been accused of repressing Uyghurs and other Muslim minority groups, particularly in the country’s western Xinjiang region.
The US State Department estimates that since 2015, as many as two million mostly Muslim Uyghurs and other Turkic minorities have been imprisoned in enormous re-education camps in Xinjiang. Beijing has long defended the crackdown in Xinjiang as necessary to tackle extremism and terrorism.
While Alibaba said that “racial or ethnic discrimination or profiling in any form violates Alibaba’s policies and values,” it declined to comment on whether any employees involved in the project faced disciplinary action.
The company also declined to comment on how the system could have been tested without official knowledge or approval from Alibaba.
Though Alibaba insists the technology was not rolled out commercially, the company explicitly touted it to customers on a website promoting its cloud services, according to The New York Times.
“Alibaba’s website for its cloud computing business showed how clients could use its software to detect the faces of [Uyghurs] and other ethnic minorities within images and videos, according to pages on the site,” the Times wrote.
When the Times questioned Alibaba about the matter, the tech firm “edited its website to remove the references,” according to the newspaper. Alibaba declined to comment on the matter.
Alibaba, one of China’s most valuable corporations, has grown to become the world’s third biggest cloud services provider, according to research firm Gartner.
The Hangzhou-based firm is the latest tech giant facing scrutiny over the potential surveillance of Uyghurs. Last week, Huawei faced similar controversy after IPVM accused the Chinese smartphone maker of testing similar technologies.
In a joint investigation with The Washington Post, the outlet reported that Huawei had tried adding so-called “Uyghur alarms” to its facial recognition software, which would identify members of the minority group and then potentially alert the police.
Huawei later said it was investigating the issue, though it denied working to “develop or sell systems that identify people by their ethnic group.”
“We take the allegations in the Washington Post’s article very seriously,” the company said in a statement to CNN Business. “We have not and will never support the use of technology to discriminate against vulnerable or marginalized groups.”