A new study highlights a hidden challenge of using AI in medical imaging research — the phenomenon of highly accurate yet potentially misleading results known as "shortcut learning." The researchers analyzed thousands of knee X-rays and found that AI models can "predict" unrelated and implausible traits such as whether patients abstained from eating refried beans or beer. While these predictions have no medical basis, the models achieved high levels of accuracy by exploiting subtle and unintended patterns in the data.

The research article draws from interviews with key NGOs working on statelessness and children’s rights to explore whether ISIS-associated children should be treated as victims or security threats. It also examines the impact that the UK government’s failure to repatriate British children has for children’s rights and what a children’s rights approach would mean for affected families and the international community. 

The study, ‘Language in job advertisements and the reproduction of labour force gender and racial segregation’, is published in PNAS Nexus – an official journal of the National Academy of Sciences of the United States of America.

The findings provide a labour-market-wide audit of how gender/EDI language in job ads helps shape workforce gender/racial composition, as well as how labour force gender/racial composition influences gender/EDI language in job ads.