I now regularly get emails from the Virginia Bar Association’s “Senior” section which I guess makes sense since my Virginia Board of Bar Examiners license is dated in October 1988. So perhaps I’ve reached curmudgeon status and I’m just yelling at young lawyers to “get off my lawn.”
With these qualifications out of the way, some quick history. When I started practicing law the secretaries at the firm where I worked had IBM Selectric typewriters that stored about a paragraph of information. We had a word processing department where longer items were typed on PC’s by a staff that either read from legal pads or listened to those tiny dictation cassette tapes. Nobody was connected to anybody else. The peak of technology was Westlaw or Lexis searches. None of our lawyers had a stand alone PC on their desk. Fax machines had just come into vogue (we quickly learned you had to copy faxes or they would fade over time). This was “state of the art.” Within seven years (1995ish), I was at a job with a PC on my desk and we were all connected to each other. Outlook was standard (I loved the calendar especially but still kept my paper calendars for a while as back up). Obviously, email was becoming ubiquitous and, of course, by late 1999 we were all worried about the technological Armageddon that we might be facing with Y2K.
My point is change always happens and I’m not a Luddite (although in my more retro moments a small cabin in the mountains with plenty of books and a nearby trout stream sounds inviting). Yet . . . a recent report from Thompson Reuters tells me “the future isn’t just about whether organization should be adopting AI – it’s about how they can do so strategically to get the most benefit from advanced technology.” Well, maybe, but can we ease off the gas pedal a moment? Thompson Reuters and other similar ventures have a financial stake in pushing AI. This is not to say their motives are purely mercenary but let’s not pretend they are exactly philanthropic, either. Each outfit pushing the use of AI has a stake in getting lawyers to jump on board while proclaiming they are the firstest with the mostest at just $99.99 per user per month for the next forever number of years. Another quote from Thompson Reuters: “the most successful organizations are those deploying AI with a coherent strategic framework – one that acknowledges today’s challenges while building resilience for tomorrow’s competitive landscape.” Oh, really? I recently read a legal memorandum which shouted “AI” due to various features, for instance, improperly citing “not to be published” cases, and generally sounding banal and dull. We can’t all be Ernest Hemingway (choose your favorite author) but a little personality wouldn’t kill our writing and improperly citing case law is first year associate level nonsense.
Generative AI isn’t merely a new way to communicate or a new way to obtain copies of cases and statutes or a new way to store information. It’s a way analyze and organize and create legal documents. As someone involved (part-time) in education for the past 20 plus years at the law school, college, and high school levels, I can assure you AI is making unfortunate inroads on student behavior. It’s an all too easy shortcut, especially in the get-it-done now society, to just let some form of AI do the heavy lifting. Pushing this into the legal world as must have or you’ll be left behind sounds much more like marketing and less like sound advice.
When faxes started dotting the landscape, suddenly clients believed that sending a 38 page document at 5:14 pm on a Friday afternoon somehow meant by 6:00 pm it would be completely and fully analyzed and a return letter making all the necessary changes, additions, and suggestions would be sitting on their fax machine. Oddly enough, I don’t recall my brain being able to read, digest, and effectively review documents any more quickly after fax machines than before them. AI promotes a similar but more disconcerting view that somehow plugging in a query will offer up acceptable answers on a virtually immediate basis. I know that it will work with remarkable speed, accuracy, and precision with good queries but I also know from doing on-line legal research for several decades that a slight modification in a query can make all the difference. If lawyers assume the first query does the trick, will they be lulled into thinking this is sufficient? But what if it isn’t? I get this runs into the Bee Watcher Watcher problem of Dr. Suess fame[1] but it does raise the question about how often lawyers will need to check AI results. At what point do AI results stand as sufficient? How can we be sure? How many queries meet best practices standards? (See, I do know some current lingo despite my advanced age).
Yes, one can point to computer assisted research as a forerunner of AI and that would be true. Of course, the obvious difference is the attorney still reads the cases and statutes and regulations that get downloaded. The attorney still has to analyze what the cases and statutes and regulations mean given the facts and circumstances of the situation. As needed, the attorney still has to write some kind of document commenting on the impact of cases, statutes, and regulations on the situation at hand. If AI is doing this, and we can’t pretend this won’t happen, perhaps with great frequency, does practicing law ultimately devolve into practicing queries for AI? Yes, this picture sounds a bit dismal and Brave New Worldish, but we can’t ignore the push being made to get everyone on board and on board sooner rather than later.
Am I against AI? If by that you mean am I against the current rush to jump headlong into AI without carefully weighing its use from all angles, then yes. If by against AI you mean I see no value in it at all, then no. I have used Chat GPT, for instance, to try to acclimate myself to AI and what it looks like. On the whole, my experience has been moderately interesting and partially successful but certainly not transformative.
If, as we lawyers really do stand in the gap for clients, making sure justice is done, wrongs are righted, language is used effectively and legally, and provide clients excellent legal support, then shouldn’t we at least take measured steps regarding generative AI? Shouldn’t there be an air of caution and a willingness to put our foot on the brake for a moment? Shouldn’t we be assuring ourselves precisely how well it truly works without relying on marketing from those who gain the most from its use?
Will AI be transformative? Maybe it will. For right now, I will continue to say to the AI pushers: I can cut my own grass for now, thank you.
[1] https://www.schooleverywhere-elquds.com/lib/uploadbook/160841479105326Did_I_Ever_Tell_You_How_Lucky_You_Are%20%281%29.pdf accessed 8-25-2025 (p. 14 – 15) in case you want to remind yourself.
