In July of 2019, I published an article predicting AI would revolutionize the practice by automating tasks like doc review and providing initial drafts of contracts. Millions of billable hours, doing menial tasks would be eliminated. My colleagues met this with skepticism and a resounding “Meh”. I received snark and feedback from attorneys who made clear that the majority were surprised my editor even published such nonsense because they “knew” AI would never be a big deal.
They were wrong.
Within a few months of Chat GPT’s release, I interviewed the founders of multiple legal-specific AI platforms and published a follow up article predicting that a “Tidal Wave of AI” was coming at the legal profession. This time AI wasn’t going to just do doc review or other menial tasks – it was going to be game-changing. This new breed of AI was definitely going to wipe out millions of billable hours.
Again, many attorneys I knew told me I was grossly overestimating the import of this new novelty. After all, this so-called Tech Wunderkind had already hallucinated and humiliated itself into briefs that cited fictitious cases! My so-called “tidal wave” would never overcome the inside knowledge of an attorney who had litigated before the same judge, dozens of times for decades.
There are still attorneys who believe that they can continue to practice as they always have, and that AI is still outside the mainstream. They are wrong.
From PI to AmLaw – AI Goes Mainstream
While AI and especially the latest iteration of agentic AI still have some room for improvement, it is no longer the new kid at the firm; it’s part of the executive committee. From Bethesda to Boston, from seven-attorney shops to sprawling AmLaw behemoths, artificial intelligence has shifted from novel to necessary. And as predicted, its adoption wasn’t just inevitable, it was unstoppable.
Today, it’s not about if your firm uses AI. It’s about how, how often, and how much of your day-to-day minutia and management it’s performing.
Seann Malloy, the managing partner of a seven attorney personal injury firm in Bethesda, says Pre/Dicta has transformed their litigation approach.
“We’ve cut prep time by 40%,” he noted.
By analyzing behavioral and contextual judicial data, Malloy’s team now walks into court with an 85% confidence read on how the judge will rule. AI hasn’t replaced judgment—it’s refined it. “Use it to inform, not decide,” he advises.
When I spoke with Pre/Discta’s Founder, Dan Rabinowitz, he explained how Pre/Dicta works.
“Pre/Dicta has a proprietary algorithm that analyzes the case at hand and 20 years of federal case data and data on the assigned judge to determine how they’ll rule – including past decisions, net worth, political affiliation, work experience, law school attended, and much more. With these points, Pre/Dicta can deliver an instant prediction with an 86% accuracy rate.”
Think about how many outside consultants, and hours of research it would take to determine whether a judge would rule to strike a motion in advance. Now it’s done with a stroke on a keypad.
It’s Early, So There’s Still Room for Improvement
Star Kashman is a cybersecurity lawyer and managing partner at the Cyber Star Legal Group. She has been using Lexis Nexus’ AI platform. While Lexis Nexus is one of the biggest brands in the profession, it’s AI hasn’t always performed well for her.
“I have mainly used it for legal research. Lexis AI is struggling to read and interpret its own cases. Meaning when a question is asked about if there is case law with certain outcomes, or showing certain ideas, it will hyperlink to case law while vastly misinterpreting the case law and what it meant. Out of many general AI tools I have used, it seems to be pretty far behind.”
Validation from The Big Guns
However, if you consider Fennemore Craig, an AmLaw powerhouse that began its algorithmic ascent back in 2022 with “Fennemore Labs,” it’s obvious that all things AI are accelerating.
David A. McCarville, director at Fennemore Labs, recounts their proactive pursuit of innovation: “We started Fennemore Labs in 2022 with our focus on how we could use innovation and AI to better serve our clients. Our first AI application was Casetext / CoCounsel, which we used to compare how legal research with AI works vs. traditional legal research.”
Fennemore didn’t just dip a toe in the water; they went for a full semiconductor swim. The results? Clients are not merely tolerating this technological embrace; they’re cheering its arrival.
As McCarville confirms, “Our fee agreements all contain language about our use of AI and individually attorneys have discussions with clients on AI.”
This isn’t just about back-office brilliance; it’s a client-facing commitment to cutting-edge competence. Today, Fennemore Craig’s AI arsenal includes Enterprise ChatGPT, Dioptra, TrueLaw and CoCounsel. The deployment, as McCarville notes, “depends on the attorney, type of matter and client.”
This bespoke approach highlights AI’s flexibility across a plethora of practice areas: document drafting and summarization, legal research, Q&A assistance, contract review and analysis, due diligence, compliance management, analytics and insight, litigation support, e-discovery, and case assessment. Essentially, if it involves data, documents, or strategic decisions, AI is lending an algorithmic advantage.
In-house counsel at companies ranging from startups to Fortune 500s are also embracing AI and making it mainstream.
Sheldon Silverman, a senior director of legal at Fortune 500 AMD shares this humorous anecdote: “There is a joke in my organization, that I think Luminance (An agentic, legal-specific AI platform) is magic. And every time we are faced with an opportunity or a challenge, I say, well why can’t Luminance do that?”
According to Luminance GC Harry Borovick, Luminance does much more than that. In addition to its generative AI, the platform has added an “agentic AI” component that actually performs tasks that range from seconds to years, to complete and maintain!
“Lumi is our agentic AI, meaning it learns from past interactions with individual users, then tailors its actions and responses over time,” Borovick says. “For example, if our marketing team is consistently negotiating and amending the same term in a contract, Lumi will automatically notify my legal team about this inefficiency, and proactively update our standards and templates to ensure smoother negotiations moving forward. Obviously we review and approve these changes, so the legal team always stays in control.
“Lumi is smart enough to recognize when it doesn’t know the answer to a certain question, and will hand the work back to me to make the call,” Borovick adds.
As with everything, it’s important to be objective. As Kashman pointed out, there are still problems and glitches to deal with. On the other hand, if Fennemore Craig has informed their clients that they’re using a whole suite of AI, then we can consider this akin to the early stages of e-discovery: Lots of skepticism but the inevitable is obvious. AI will become a mainstream part of the legal profession within another year or two.
