Efficiency, Ethics, and Algorithms: The Implications of AI on the Legal Profession and the ABA Model Rules

Abstract

The rise of new AI language models has led to a wave of excitement and trepidation about how such tools might change the way businesses operate. With their remarkable ability to create polished content, AI’s potential to transform the service economy is self-evident. This potential will also impact the legal profession. Three possible sectors include legal drafting, legal research, and decision making and recommendations. AI tools on the market now are billed as being able to handle these tasks, eliminating countless labor-intensive hours from firms. These tools can transform the way lawyers practice by performing legal research and generating fully written legal motions and briefs, and could even make recommendations in legal matters. This shift presents a number of ethical concerns: malpractice concerns, if lawyers abdicate a significant part of their work to AI assistants; due process concerns, if AI tools take over tasks in the criminal justice system; and labor concerns, AI assistants take over the majority of work currently performed by staff.

Some of these concerns are future-focused, and others involve tools widely available right now.

One current popular tool is ChatGPT. While there are a few legal-specific AI tools available, they are not yet widely available; ChatGPT is relatively inexpensive, and in some cases, free. ChatGPT has the potential to help attorneys work more efficiently and effectively. And yet the ABA Model Rules are silent on how — or even whether — to use ChatGPT. The second half of this paper highlights common scenarios in which an attorney might use ChatGPT, and which Model Rules apply. In doing so, the paper provides speaks to our current moment, one in which lawyers are considering a new tool, and are unsure of what potential ethics rules implications might be. Ultimately, whether an attorney uses a law-specific AI tool, ChatGPT, or something else entirely, they should proceed cautiously and always with client consent.

Keywords: AI, artificial intelligence, ChatGPT, CoCounsel, Harvey, risk assessment, malpractice, due process, client confidentiality, Model Rules, ethics, technology