Is Australia Ready for AI on the Bench

The rapidly-accelerating integration of artificial intelligence into our lives will soon affect court rooms and other legal environments in Australia. This will occur so quickly and so soon that critical threshold issues for a smooth integration of AI into court environments remain unexamined. In particular, the psychological- and attitudinal-readiness of the judiciary to work with algorithmic tools is in question. Further, the literature is silent as to the degree to which judicial decision-makers are open or resistant to the fettering of their discretion by inscrutable AI systems. This distinct gap means that we do not yet know (1) whether Australian judges understand AI, (2) whether they trust it, (3) whether they appreciate its potentials and risks or what might influence their attitudes, (4) whether they are aware of the differences between automated and augmented decision making, and (5) what accountability and oversight mechanisms will be required. This paper considers the likely barriers and risks to a successful integration of AI into the work of judges in Australia, based on research and experience in other jurisdictions – mainly the United States, where algorithms already play a significant role in facilitating judicial-decision making.