Zach Cass opened the CLOC Global Institute this week with a question: if you could automate everything, what would you leave in? It’s a critical question for legal ops and for the legal profession .

I’ve been told C-suite executives are already wondering how to build agentic workflows that reduce the need to consult in-house counsel. Asking outside counsel for something will be a last, last resort.

So what’s left? What do we leave in? Where will human lawyers still add value? Hard questions.

Here’s my post on the keynote and the questions for Above the Law.

I recently attended another legal tech conference celebrity keynote that left the audience wondering “what does this have to do with me?” Here’s my post for Above the Law on why celebrity keynotes at legal tech conferences sometime miss the mark. And what conference organizers and, for that matter, the celebrity speakers should be focusing on: application of the presentation to legal, making the talk relevant to what we do everyday, and not talking about being a celebrity for the sake of being a celebrity.

Legal ops’ biggest annual gathering, CLOC’s Global Institute, is moving to McCormick in Chicago this week and, as can be expected, AI is shaping the agenda. One big question: how will AI reshape legal ops in particular, and how do its practitioners prepare? Not to mention how attendees will deal with the new venue (we aren’t in Las Vegas anymore!). Here’s my preview for Above the Law.

The Smokeball/Thomson Reuters partnership is the latest effort to better tie the legal research and substantive side of the business to the administrative side. What’s unique about it is that TR and Smokeball both say their partnership is purposefully directed toward the smaller law firms. It’s particularly timely as smaller law firms are struggling to compete with bigger law firms with more resources.

My post for Above the Law.

Another week. Another law firm caught citing cases that don’t exist. But this time it was Sullivan and Cromwell, one of the most influential firms in the world. 36 errors. Three pages to describe them. Fabricated passages from real cases. S&C said its AI policies weren’t followed. That it had training designed to prevent exactly this.

We can keep talking about education, policies, and fines. We’re doing all of that. But it doesn’t seem to be working. Moreover, it isn’t just a lazy lawyering problem anymore. When the public hears that lawyers are citing cases that don’t exist, their first instinct, as I discovered recently, is that the lawyer made them up to win. Of course, 

that’s not what typicaly happened. But the fact that it’s the immediate reaction says something about where we are. As Judge Bernes Aldana put it at a recent ABA conference: “The legitimacy of our courts depends on the public’s trust and confidence.” That’s why AI hallucinations aren’t just a sloppy practice problem. They’re helping to erode of the rule of and trust in law itself.

What can we actually do about it? Stronger penalities? Bar discipline? Better education and awareness? Maybe all three.

My post for Above the Law.

Just returned from ILTA’s Evolve 2026 conferance in Denver and here’s my Above the Law post. Three days of content at a breathless pace. A record 500+ attendees that tested but didn’t break the small-conference vibe that makes Evolve worth it. An opening keynote from Zach Abramowitz that framed the AI moment as well as anything I’ve heard on a conference stage this year. And a vendor experience that reminded me that hallway proximity often beats cavernous exhibit halls. Once again, ILTA got most things right. 

I will have to say, however, the AI content conversation in legal tech needs to grow up a bit. We should be past simple prompt writing exercises and talking about the repetitive tasks AI can do. Addressing critical questions like where is GenAI taking the profession, and what does that mean for long-range planning are still largely missing from most conferance agendas.

Evolve’s cyber content, on the other hand, was exactly right: technical, practical, and even emotionally honest about what a data breach actually does to the people who have to manage and live through one.

I do hope ILTA doesn’t let success turn Evolve into a spring version of its sprawling late-summer flagship. The small conferance mission is still the magic. 

Law firms are panic buying AI to satisfy client demands and it’s backfiring. Clients are making demands that their firms get AI but often don’t know what they really want. Firms don’t know what they need. It ends up being a hot mess of wasted money, unused tools, and unhappy clients. It’s a classic perish for lack of knowledge. The solution starts with education. Not more AI slop.

Here’s my post for Above the Law. 

Here’s my Above the Law post on Zach Abramowitz’s keynote at ILTA’s Evolve conferance. The keynote lived up to its title: Most Law Firms Are Doing AI Wrong. Here’s How to Do It Right.

His argument is that the failures in GenAI adoption from ineffective training to analysis paralysis to hallucination panic, all trace back to the same root cause. Firms are deploying AI without really understanding it. And you can’t use what you don’t understand.

Some key observations:

Hallucinations aren’t a glitch. They’re a feature of how GenAI works. Understanding that changes how you deploy it.

Stop asking what GenAI can do. Start asking what you ought to be doing because of it. That’s a fundamentally different question.

AI ROI isn’t efficiency. It’s the improved results, expanded capabilities, and competitive differentiation it can bring.