Editor’s Note: The author of this post is a partner at Bryan Cave.
By Kathryn DeBord, Chief Innovation Officer, Bryan Cave LLP
Recent headlines about Artificial Intelligence (AI) flooding legal press in recent months with accompanying images of human-like robotic associates have many lawyers asking: will AI really take our jobs? And the follow up: can Siri or Alexa help me quickly research basic legal principles?
My short answers are currently “no,” and “not quite yet.” The answer to the former is not likely to change; the answer to the latter is subject to change at any moment. In the meantime, my discussions around AI, or cognitive computing, and its place in the law, as well as my recent transfer from litigation partner to innovation partner, have allowed me to reach some (preliminary) observations about it all, and preview where I think the technology can and should be going in the practice of law – particularly, for the purposes of this article, litigation.
At root, AI processes a lot of information and identifies patterns within that information. And the more humans interact with it, the more it can learn, suggest, and identify new patterns. Right now, AI is used in the legal field primarily to facilitate document review and to conduct due diligence as well as legal research. But the use cases for AI have as much possibility as we can imagine them.
As I imagine the possibilities for AI, I draw on my history as a commercial litigator. A litigator’s job, at root, is to show that his or her client should win under the applicable law. To do that, they need to (1) identify which legal principles apply to their case, or should apply to their case; and (2) develop the facts of their case in a way that is consistent with those legal principles.
Identifying the legal principles that are applicable to a particular set of facts is not clear cut though, particularly in complex commercial cases. More often than not, the law that is applicable to any given case is itself up for debate (and litigation before the court). Indeed, our entire body of common law is effectively created through litigation and advocacy, which ultimately creates new law. And when the law that is applicable to a specific fact pattern is not clear cut (and must therefore essentially be “created” based on existing legal precedent), a litigator looks for patterns in that existing precedent. Litigators analyze cases and statutes to try to draw analogies between their fact pattern and the fact patterns that existed in previously decided cases, and the analogies that litigators look for are confined to one area of law or one body of law.
The possibilities for AI in this context are awesome, and entirely feasible. AI could ingest a statement of facts written by a lawyer to accompany a summary judgment brief, and produce case law decided based on similarly patterned facts. AI could identify patterns and similarities across different statutes to augment statutory construction research, or across case law and other legal resources to analogize legal principles. The application of antitrust laws can be analogized to other federal consumer protection policies identified by AI.
AI could suggest, through similar patterns in the Bankruptcy Code and statutes, that the power of an intervenor in a receivership proceeding can be analogized to the power granted to a debtor in possession under the bankruptcy code. Lawyers could argue the proper interpretation of the Fair Credit Reporting Act – which is infamously vague – by analogizing it to other federal statutes, like the Telephone Records and Privacy Act, with similar statutory language or cases interpreting that language that are identified by AI.
Today, the ability to identify, articulate, and argue patterns and analogies, even in instances where the analogies are not immediately apparent or cross areas and bodies of law, oftentimes distinguishes the good litigator from the average and can become the deciding factor in the case. If AI augments this process, it could not only supplement a litigator’s human analysis and judgment, it could aid the courts confronted with creating new law, and ultimately have a hand in the creation of new laws. That is powerful, and maybe a little eerie.
Eeriness aside, as I think about the possibilities for AI, the concept of “Augmented Intelligence” may actually be more apt in a legal context. Augmented Intelligence generally implies supplementing human thinking versus reproducing it. instead of “Artificial Intelligence.” To be sure, even as it works for us to help us evolve our laws, AI will never replace the multi-layered analysis and decision-making that good lawyers engage in every day to represent their clients well. If AI is thoughtfully developed in the legal space, sooner than we think, Siri (or Alexa) may be able to articulate not just basic legal standards, but also suggest analogous legal principles they would never had thought of on their own to help litigators litigate and ultimately create new laws.