(Artist’s depiction of exhausted articling student, circa 2020.)
I have no idea but the lawyer in me says, it depends. (Thank you very much, I’m here all week).
This heavily researched and sufficiently cheeky article looks at the question of the unauthorized practice of law and asks how that reconciles with AI making and executing legal decisions in the practice of law. The problem arises that in order to practice law, one has to be a licenced individual, and the then question becomes whether or not the AI is practicing law without a licence.
More holistically, this raises a discreet point about regulation of AI. What happens when AI moves into functions currently regulated to be performed by licensed individuals? Will it be enough to have simple oversight via the possibility of intervention in the AI, as my first principle suggests? Or will there have to be continuous engagement by human monitoring? Or are we on the road to moving away from licensure of individuals in respect of the job to be performed?
What happens to doctrines like legitimate expectation of procedural fairness if certain administrative procedures are taken over by AI? How does one test if promised procedures have been adhered to if there is no person to explain how an administrative body applied its actions when previously officers of an administrative body could have been called to account?
Administrative bodies are going to need to deal with these concepts in advance or, as the article points out, they are likely to end up dealing with them in arrears and run the risk of having outcomes not always favourable to their members.
A robot and a lawyer walk into a bar.
The bartender looks up and says “Get the f**k outta here!”