Commentary on Political Economy

Wednesday, 9 December 2020



Xinjiang data platform deems people suspicious by association

The ‘high-tech’ programme relies on local authorities inputting data and carrying out interrogations © AP

Christian Shepherd in Beijing

A big data platform helped Chinese authorities in Xinjiang to identify individuals deemed suspicious by association that facilitated mass detention of Muslims in the region, a leaked government document has shown. 

Over the past four years, more than 1m Uighurs, Kazakhs and other mostly Muslim peoples in the far western region have been detained in extrajudicial camps where they are subject to political “re-education.”

Beijing defends the programme as necessary to eradicate “extremism” and claims the detainees are students who volunteered to undergo training in the facilities. 

Predictive policing platforms are really just a pseudo-scientific fig leaf for the Chinese government to justify vast repression of Turkic Muslims

Maya Wang, Human Rights Watch

But a document generated by a police database known as the “integrated joint operations platform”, or IJOP, shows how mass surveillance aided and encouraged authorities to detain individuals for everyday, lawful and non-violent behaviour, according to Human Rights Watch, the advocacy organisation that analysed the document.

The group said that the list of 2,000 individuals from Aksu, a predominantly Uighur prefecture in southern Xinjiang, appeared to be created by identifying individuals who had talked to, or were in some way related to, an individual deemed “suspicious”.

Local authorities then used behaviours such as phoning a relative overseas, reading a Koran without state permission or using “suspicious” smartphone applications including Skype to determine who to detain. Some information was also supplied by the IJOP system.

The Aksu list’s contents and use, as described by Human Rights Watch, closely match those of another leaked document drawn up by local officials in Karakax county, in nearby Hotan prefecture, and vetted by the Financial Times.

“‘Predictive policing’ platforms are really just a pseudoscientific fig leaf for the Chinese government to justify vast repression of Turkic Muslims,” Maya Wang, a researcher at Human Rights Watch, said.

The document was sent in late 2018 from an anonymous source in Xinjiang to Radio Free Asia, a US state-funded broadcaster. The list was passed on to Human Rights Watch in August.

The organisation matched phone numbers and other details in the document to individuals in Aksu to verify its authenticity. The FT was unable to confirm independently the group’s assessment because the document was not publicly available.

Human Rights Watch first unveiled China’s use of the system in 2018 and later “reverse engineered” the application to show how it was used by law enforcement to gather data and find suspects.

Although billed as a high-tech solution for what Beijing claims are security threats in the region, the technology remains largely reliant on the daily grind of the region’s law enforcement, who input data and carry out interrogations based on names the system provides.

Once flagged as “suspicious” by the system, an individual will probably be singled out by other government agencies as the platform is increasingly being integrated into screening processes, such as for police jobs or public services, Human Rights Watch said. 

Aksu’s Public Security Bureau did not respond to a request for comment.

No comments:

Post a Comment