Agentic AI and Legal Services
When ChatGPT first dropped, the feeling was borderline magic. You'd type something in and this thing would respond with what felt like a human being. But even then, I always wondered: what if it could actually do things? Not just talk about them. What if it could reach into your calendar, your inbox, your systems, and actually execute?
That question is starting to get answered. Last weekend I spent some time playing around with a tool called OpenClaw. For those who haven't come across it yet, it's an agentic AI platform. Meaning it doesn't just answer questions. It actually controls software, takes actions, and executes tasks the way a human would.
And the implications for legal and professional services are enormous.
What's Actually Changing
I came across a podcast recently featuring Jordan Furlong, a legal sector analyst and forecaster based in Ottawa. He's been writing extensively about where the legal profession is heading, and his thinking is sharp.
His core argument is this: the means of production in law are shifting. For decades, law firms made money the same way. Hire lawyers, bill hours, create leverage, distribute profits. The value was in the human effort. But that model is breaking down. AI is increasingly becoming the conduit for legal work. The workflow gets operated by an intelligent system. The lawyer's job becomes supervising that system, validating its outputs, and exercising judgment.
He argues that competing on expertise alone is a short-term play. Within a few years, the smartest lawyer in the world will be an AI, and it won't be close. What you compete on instead is who you are. Your character. Your integrity. Your ability to sit with a client in a moment of uncertainty and give them a clear, trustworthy call on what to do next. That's the thing machines can't replicate.
I think he's right. And playing around with OpenClaw this weekend made it feel a lot less theoretical.
My Own Take
So where does that leave us? Here's what I'm sitting with.
1. Security and privacy are the first wall to climb.
There's a big difference between letting an AI assistant schedule your meetings and send Zoom links, and letting it handle legal work. Client confidentiality, privilege, data sovereignty, professional indemnity. These aren't minor concerns. They're foundational. Before agentic AI gets anywhere near a legal file, the industry needs serious frameworks around how that data is stored, accessed, and protected. We're not there yet. But the pressure to get there is building fast.
2. Uptake will be slow, then sudden.
This is how most transformative technology works. A long runway of early adopters, sceptics, and incremental progress. Then a tipping point. Then everyone's using it and the people who didn't prepare are scrambling. My read is that we're probably anywhere between 5 to 20 years from that critical mass moment in legal. But the firms that start building the muscle now will be the ones who own the transition when it comes.
3. Boutique firms are uniquely placed to move first.
This is the part I find genuinely exciting. We're small. We're nimble. We don't have 47 layers of approval to get through before we can try something new. A five-person firm can adopt a new tool, test it on real work, iterate, and build a real competitive edge before a large firm has even finished its internal review process. The speedboat advantage is real here. If you've been reading these emails for a while, you know I believe boutique firms can punch well above their weight. Agentic AI is one of the clearest examples of why.
4. Guardrails and AI literacy are non-negotiable.
None of this works without people who understand how to use these tools responsibly. Not just technically, but ethically. Knowing when to trust the output and when to question it. Understanding the limits of what the system can and can't do. Building the habit of human oversight rather than blind delegation. If you want a starting point on this, I'd point you toward a company called 43 Degrees Below, who do excellent work in AI literacy training. Worth a look.
The Bigger Picture
What struck me most about Furlong's thinking, and what my weekend with OpenClaw reinforced, is that the question isn't really whether AI will change legal. It already is. The question is whether you're going to be someone who shapes how that change lands in your firm, or someone who reacts to it after the fact.
The lawyers who thrive in this next chapter won't necessarily be the most technically brilliant. They'll be the ones who are trusted. Who show up with integrity. Who can take a client from uncertainty to clarity and charge accordingly for that journey. The same traits, when you think about it, that have always separated the great lawyers from the rest.
The tools are getting better fast. The opportunity is real. But the foundation is still the same thing it's always been. Who you are and whether people can count on you.
More next Tuesday
Gordon