The organisation’s chief security team uses facial recognition, voice recognition and even track typing patterns to look for inconsistencies in the way candidates take their exams online, to arrest fraudulent practices.