When we allow AI to make critical decisions about individuals, who gets to scrutinise them? Independent auditors struggle to investigate harmful impacts of algorithms due to strong barriers to data access. Our latest research suggests a path forward to reconcile AI accountability with data privacy protection.
