By Kirsty Needham
SYDNEY (Reuters) – An Australian university said on Wednesday that it has asked journal publisher Wiley Online Library to retract a research study conducted by a former faculty member that helps Chinese facial recognition software better identify ethnic Uyghurs.
Curtin University said that the study published by the academic Wanquan Liu and co-authors at Chinese universities in 2018 was conducted without its knowledge and had breached its ethics code.
“Ethics approval should have been sought by the academic,” the university said in a statement to Reuters in response to a request for comment on the outcome of its internal investigation into the matter.
Liu did not immediately respond to an email request for comment.
Wiley issued an expression of concern to readers in September 2020 about the data contained in the study, and told ABC, Australia’s national broadcaster, on Wednesday that it was reviewing Curtin’s request for a retraction. Wiley did not immediately respond to a Reuters request for comment.
Liu and his co-authors said they had collected a dataset of facial images by recruiting hundreds of students of Uyghur, Tibetan and Korean ethnicity from the Dalian Minzu University in China.
The study, funded by the Chinese government, said “face recognition has great application potential in border control, customs check, and public security”.
The Australian Broadcasting Corporation’s Four Corners programme first reported in 2019 on concerns that the Uyghur subjects had not given consent.
Liu worked at Curtin in Western Australia for two decades, funded by an Australian Research Council grant, but left to take up a new post at Sun Yat-sen University in Shenzhen in May, according to biographical details published on the Chinese university’s website.
The incident is the latest in a spate of controversies involving Chinese ethnic profiling research on Uyghurs, a mostly Muslim minority group under heavy surveillance by authorities in the western Chinese region of Xinjiang.
China denies all accusations of mistreatment or discrimination in Xinjiang.
Another publisher, Springer Nature, has retracted two articles involving Uyghurs on ethics grounds in the past two weeks.
A Curtin University spokesperson told Reuters in an emailed statement the research had been undertaken by Liu informally, without the support of a contract or formal documentation, though the published study acknowledges him as working in Curtin’s computer studies department.
Australia’s universities have come under government scrutiny for foreign interference through international research collaboration. The Australian Research Council said in March that security agencies had increased vetting of projects funded by its grants at the country’s universities.
(Reporting by Kirsty Needham; Editing by Simon Cameron-Moore)
