Last week I talked about how legal AI has changed beyond recognition. This week I want to get specific about what that actually looked like in practice.

A huge part of my life as a disputes lawyer was drafting. Correspondence, pleadings, applications. And the reality of drafting, the bit no one talks about, is how much of it isn't really writing at all. It's leafing through the underlying documents trying to find the one paragraph you half remember. It's pulling up a precedent and reworking it line by line. It's writing a sentence, deciding it's not quite right, rewriting it, checking it against the source material, and rewriting it again.

By the time you've done all of that, the deadline is breathing down your neck and the part of the job that actually matters — the judgment calls, the strategy, the structure of the argument — gets whatever time is left over.

When I brought Legora into that process, the way I got to the end goal completely changed. Instead of spending hours cross-referencing materials and fishing out the relevant details myself, I could feed in the underlying documents and get back the key points distilled clearly, ready to work with. That meant the time I used to spend wrestling text into shape was spent on substance instead, considering the draft properly, reworking it, applying my own judgement.

That's what I mean when I say the tools have caught up. They haven't replaced the work. They've made space for the part of it that matters most.

I'd argue that lawyers who aren't using AI are already behind on their ethical obligations.

Every lawyer has a duty to stay competent. That means keeping up with changes in the law, but also changes in HOW law is practiced.

AI is a change in how law is practiced.

If you can use it to work faster, research more thoroughly, and serve your clients better, and you're choosing not to - I think that's worth examining against your competence obligations.

I use it in my own practice - forms, information gathering, parsing documents quickly. It saves time. That time saved gets passed to clients.

But there's a line.

Last year alone, literally thousands of cases were flagged for citing law that didn't exist. It was hallucinated by AI and nobody checked. Many of those led to sanctions.

AI didn't get those lawyers sanctioned. Not checking the output did.

AI can read a statute. It can draft a demand letter that looks completely right and is missing five things a lawyer would catch. It doesn't know what it doesn't know, and neither will your client.

The legal judgment still has to come from you.

Use it and be more effective for your clients. Just don't outsource the part that actually requires a lawyer.

Most lawyers are still figuring that out. The ones who aren't are already falling behind.

New Article, possibly my last for a while.

I've spent two years figuring out how to make a two-person law firm compete with teams twenty times its size using AI. This is the closest I'll come to explaining how.

Also explains why I can type “plz fix” and get back work product that reads like I spent three hours on it, when really I spent three hundred hours building the system that did.

Yup. I use AI to brainstorm sometimes, but the last time I used AI to synthesize my thoughts into a very important letter/email draft, it still took me hours to get into the form I wanted. It is all my work in the end with just a little help with structure and pulling two different emails together into one. Probably saved me 1 hour of 5? So, helpful, but not doing my work for me

I’ve never cared for the term “prompt“ for AI.

I use Claude ai regularly, but I don’t think of what I do as prompting it. I’ll ask questions, have it ask me questions, engage it, berate it, push it to do better.

I find working with AI is an iterative process.

I think too many lawyers treat an AI model like a sycophantic search engine that has a tendency to make things up (although I find that the latest models seem to have greatly reduced hallucinations). But that misses the point. This is a computer you can “program“ by talking to it. You can hone the algorithm with just a sentence.

Legal Practice – AI empowerment not AI replacement

I read a very recent article where a Tech CEO stated “So white-collar work, where you’re sitting down at a computer, either being a lawyer or an accountant or a project manager or a marketing person – most of those tasks will be fully automated by an AI within the next 12 to 18 months.”

Working at InfoTrack I’ve come to spend more and more time using and understanding AI not least of all because we at InfoTrack are fully committed to AI and AI-enabling products for the benefit of our clients and the profession, but I am also exposed to how the profession is adopting and using AI.

Yes the rate of AI adoption in the legal industry is unprecedented, and not a day goes by when I don’t look back on a career spent trying to convince law Firms to early-adopt and wish we had seen this rate of adoption all along, but the stated narrative in the article definitely doesn’t track with what I’m seeing and experiencing…

AI is already excellent at tasks, like drafting, summarising, extracting key information, analysing large datasets and it will undoubtedly keep improving, but last time I checked the practice of law is not a collection of isolated tasks. It’s sound judgement under uncertainty, it’s risk allocation, it’s navigating ambiguity and it’s advising a clear path forward when there isn’t a clean answer to be found, and critically — it’s a lawyer’s willingness to be held accountable and stand behind their advice.

Legal work operates inside a framework of licensing, professional standards, insurance, regulation and personal liability. Clients aren’t paying for text generation or task completion, clients are paying for sound judgement and accountability. Technology can accelerate output, but it does not automatically assume responsibility – and that distinction matters.

So, I believe what’s far more realistic to expect than eradication of the profession is acceleration and augmentation. Routine work compresses, turnaround improves, capacity increases and the relative value of judgement, strategy and client counsel goes up. That doesn’t weaken the profession - it sharpens it. Big automation claims generate urgency and urgency generates uptake, but professions built on trust and regulation don’t transform overnight. They evolve deliberately — especially where accountability sits squarely with a human.
AI will reshape legal work there’s no doubt about that, but replacement is a very different claim from transformation.

I’m optimistic about AI in legal, but I’m just as optimistic about lawyers. I don’t believe the immediate future is AI replacing the profession — it’s AI empowering it. In the right hands, AI can unlock enormous upside, including greater access to the law and to legal representation.