Artificial Intelligence, the “AI Justice Challenge” and the Future of Law

 

Artificial Intelligence, the “AI Justice Challenge”  and the Future of Law

A google search of “industries disrupted by artificial intelligence” returned an article listing 12 industries: Medical Diagnosis, B2B Processing, Mathematical Analysis, Resource Scheduling, Manufacturing, Transportation, Surgery, Insurance, Education and Law.

Artificial Intelligence (“AI”) applications in law are now a reality, as they are for several broad swaths of industry in our society. The use of this technology is ostensibly aimed to improve the justice system by streamlining and simplifying – and should ultimately support greater access to justice to society. There are, however, important cautionary considerations as this technology continues to proliferate.

It’s important to recognize the use of AI is meant to be assistive, be user experience based and cast a net to improve access to justice. As a counterpoint, there is a growing and controversial use of AI to assist with decisions based on recidivism and historical data: the purpose of this application is to speed-up sentencing to reduce backlog that exists within the justice system.

For instance, in Wisconsin, a judge imposed a long sentence on a black defendant, Eric Loomis. Loomis challenged the sentence, and asked to see the considerations that went into his “criminal risk assessment” – an assessment that was generated by an AI, and used by the original presiding judge. The state Supreme Court (the appeal court) declined to hear the case. Whether the algorithm’s output provided a sufficient level of transparency, and to what degree it relied on historical data, remains unknown. The worry is this decision has disproportionately impacted dis-advantaged populations.

Similarly, California recently passed a law that requires state courts to use machine learning and other statistical tools to sift through the backgrounds of people accused of crimes. The law is intended to determine whether suspects should be incarcerated prior to trials.

Both Wisconsin and California use historic data to generate correlative insights. This approach reinforces embedded biases and continues a vicious cycle for disadvantaged populations. The data presented to a judge, essentially, becomes another witness against the user in the justice system – the transparency of which is neatly obscured in a computer algorithm. Artificial Intelligence that is used to repeat the historical mistakes of the past is a misapplication of this disruptive yet promising technology.

British Columbia’s approach to Artificial Intelligence in its justice system is being guided by the desire to create an assistive, streamlined, efficient, effective and accessible experience for citizens going through a legal process.

The justice sector partnered with Innovate BC, and BC’s technology innovation community to launch an “AI Justice Challenge.” The goal is to improve and optimize the user experience with solutions developed in five different areas:

  • Smart Online Guide: an aid to complete forms easily and correctly, respond to citizens in their preferred language and aids to converse.
  • Intelligent Reviewer: a system that can intelligently parse a document and retrieve salient and relevant points quickly.
  • Online Justice Chatbot: AI to provide legal information or guidance to matters involving probate, wills and estates planning.
  • Auto Transcriber: faster, cheaper transcription.
  • Smart Court Inquirer: an interactive platform for tracking trial processes, including an augmented reality application for way-finding.

While BC’s approach focuses on optimization of the user experience, many American jurisdictions have begun the use of AI to address efficiencies that aim to reduce recidivism, and reduce pressure on prison systems. This use of AI is very concerning. British Columbia’s focus on access to justice and user experience, conversely, is believed to be a much better use of the technology, and will ultimately be more effective in bringing the efficiencies needed in the justice sector.