Across the world, experimental technology is being deployed that promises to recover hundreds of millions of euros in alleged welfare fraud. Some of the most vulnerable members of societies are being stripped of benefits and placed under investigation after being flagged by opaque systems whose inner workings are impossible for them to question. An international team of journalists gained unprecedented access to one of these "suspicion machines" and revealed the extent of discrimination and bias encoded within. How can more journalists interrogate these predictive algorithms and the claims of accuracy and fairness being made about them? What skills and support are needed for this form of accountability reporting?

Organised in association with Lighthouse Reports.

Photos