Artificial Intelligence and the Law

Plexus, a NewLaw law firm based in Australia, has just released a new legal product that purports to apply artificial intelligence to a relatively common, discrete legal issue -- detemining whether a proposed trade promotion (advertisement in US parlance) is in compliance with applicable law. 

In the video below, Plexus Managing Partner Andrew Mellett (who is a MBA, not a lawyer), observes that this type of legal work would ordinarily take four to six weeks to complete and cost several thousand dollars.  Mellett claims that the Plexus product can provide "a legal solution in 10 minutes" at 20% to 30% of the cost of the traditional consultative method -- no lawyer required, albeit Plexus lawyers were the indispensible architects for the underlying code. 

From the video, it is unclear whether the innovation is an expert system -- akin to what Neota Logic or KM Standards are creating -- or artificial intelligence (AI) in the spirit of machine learning used in some of the best predictive coding algorithms or IBM's Watson applied to legal problems.   Back when Richard Susskind published his PhD dissertation in 1987, Expert Systems In Law, an expert system was viewed as artificial intelligence--there was no terminology to speak of because the application of technology to law was embryonic.  Now we are well past birth, as dozen of companies in the legal industry are in the toolmaking business, some living on venture or angel funding and others turning a handsome profit.

My best guess is that Plexus's new innovation is an expert system.  But frankly, the distinction does not matter very much because both expert systems and AI as applied to law are entering early toddler stage.   Of course, that suggests that those of us now working in the legal field will soon be grappling with the growth spurt of legal tech adolescence.  For law and technology, it's Detroit circa 1905.