IN THE PROFESSION
Benz Stadium. The surrounding short-term rental market could see a $ 70 million boost. And, of course, jobs. The human resource need will be incredible with huge job creation.
What about Artificial Intelligence?
It is good and bad. On November 11, 2025, the State Bar published a 16-page Artificial Intelligence guide. The guide is intended to provide“ practical, reliable, guidance” on the use of generative AI for lawyers in our practices. Here is a link to the guide: Click here.
You might be surprised to see that much of the use of AI is governed by existing provisions of the Georgia Rules of Professional Conduct( GRPC).
A summary list follows: 1.1 Competence; 1.7, 1.9, and1.10 Conflicts of interest; 1.6 Confidentiality; 3.3 Duty of Candor; 5.3 Non-lawyer Assistant; 1.5 Reasonable fee; and 5.1 and 5.3 Supervision.
The guide analyzes the foregoing rules and more in the context of AI usage. Like it or not, AI is, now, very much a part of the practice of law. From the biggest cases to the smallest.
GRPC Rule
1.1 requires all of us to maintain a level of competence. I submit to you that irrespective of the size of your firm or the cases you are working on, some knowledge of AI is now the standard of care. Don’ t believe me? Get caught submitting a brief that includes AI hallucinations, and see what happens. And that is just the beginning.
The Guide also includes a robust discussion of risk mitigation considerations. Topics include, hallucinations( bad citations and / or incorrect answers), confidentiality and training data, and security.
Our firm is small, but most of our cases involve millions of dollars. Every day, AI creeps further into our practice. And, as I wrote, above, it is both good and bad. Separate and apart from the focus of the Guide, here are some issues that we are tracking.
Your clients are asking AI for legal advice. AI is happy to oblige.
Just the other day, a new client started our conversation about the case by instructing us that she would like us to be quiet and listen to her thoughts for 10 minutes. The client’ s thoughts included a generous dose of legal terminology and concepts. The client used that information to weave an argument that almost made sense. I was thinking, Dan Brown novel. But, each argument, included dubious assumptions, omitted critical bad facts, or misapplied legal concepts. While the client was speaking, we quickly realized she had dipped deeply into the AI well. We considered how to respond without insulting this new client. When the client completed that part of the discussion, we proceeded to address the arguments.
What is the take-away from this story?
Several, I think. First, AI is literal. Ask the wrong question – get the wrong answer. Next, AI is programmed to attempt to detect what you want, and to give it to you. Think about American Idol –“ Simon, you don’ t know what you are talking about. All my friends say I am a great singer.” AI is a pleaser. It will try to tell you what it thinks you want to hear. Also, AI sounds authoritative and many of your clients and potential clients will trust the answers. Don’ t be too quick to dismiss AI answers. They may be correct and even good ideas. On the other hand, some answers will be bad. Your clients and potential clients might be inclined to“ confront” you with AI information.“ What about this?”“ Why didn’ t you think of that?” Who cares, except that it does undermine confidence and it can poison your relationship with your client. And worse. Read on.
Your clients’ AI inquiries are not privileged.
Your clients do not know that their inquiries and the responses they receive are not privileged.
Recently, I deposed a Defendant who was accused of misappropriating intellectual property he developed with his soon to be ex-wife. Husband and business partner put the business in the name of the business partner in an effort to defeat inclusion of the business asset as marital in the divorce. Husband and the business partner texted about what they termed a“ Sopranos style agreement.”
When I asked about the Sopranos style agreement in the deposition, the husband repeated a strange seemingly rehearsed answer with some buzz words that seemed unusual. Historically, I would have suspected lawyer coaching, and upon inquiry about the source of the weird answer, I would draw a privilege objection. On this occasion, it turned out the client asked an open AI platform how to answer when asked. No privilege applied and we received all we needed to amount to a damning admission. As you might imagine, the case quickly settled.
What is so great about AI?
First, open AI is great for all sorts of things. I ask questions like,“ How do I make an old fashioned? How long will it take to drive to Macon? What is the best temperature for coffee?
16 MARCH APRIL 2026