'When and How Should We Invite Artificial Intelligence Tools to Assist with the Administration of Law: A Note from America' by Jeff Ward in (2019) 93
Australian Law Journal 176 comments
Many non-technologists – lawyers and judges included – shy away from direct engagement with advanced technological design, development, and deployment. This is understandable. The details of important technologies like artificial intelligence (AI or AI tools) can seem intimidating and overwhelming. Given the emerging importance of AI in almost every aspect of our lives, however, we will be better served by widespread direct engagement, especially when AI touches our core legal institutions and processes. This is my invitation to you, and I aim to give you both encouragement and a few tools.
We are all familiar with the ways in which the machines of the Industrial Revolution supplemented and supplanted human muscle in endeavours of production. Yet, the provinces of the human mind have long seemed out of the reach of machines, and legal professionals have had to devote little attention to AI-driven disruption.
Recently, that has changed. AI is now used to accomplish innumerable tasks that would previously have required human intelligence. And modern AI tools – especially a powerful subset of AI powered by machine learning – are assisting with sophisticated research, decision-making, judgment, and prediction in ways that are both indirectly and directly meaningful for the law. The law is affected indirectly, as people and organisations now exist in an increasingly AI-rich world, with AI deployed in finance, medicine, education, transportation, and much more. The law is also affected directly, as a growing number of practitioners are using AI for a broad range of legal tasks, such as legal research assistance, technology-assisted review of documents to determine relevant and discoverable information, and case selection and valuation. And in the courts, AI is influencing evidentiary determinations, pre-trial bail and parole decisions, and even sentencing, the last of which will be explored below.
Yet, this rapid growth in the use of AI is neither pre-determined nor inevitable. Before AI tools outrace our professional judgments, we must ask: When and how should we invite artificial intelligence tools to assist with the administration of law? To be sure, the care and engagement of legal professionals in the design, development, and deployment of AI-driven tools should grow commensurately with the extent to which fairness and due process concerns are at stake.
In this short article, I offer: (1) a brief overview of the pros and cons of bringing AI into the law, (2) a brief, well-known case study from US law on AI in the context of criminal sentencing to highlight the need for legal professionals to remain engaged, and (3) a basic AI explainer along with a line of questions that non-technologists can ask to engage with AI in various contexts.
