Lawyers, AI, and the Power of Deliberate Design: What Happens When We Stay in the Loop?

Not just faster. Not just cheaper. But better.

That’s the headline coming out of a recent randomized controlled trial of cutting-edge legal AI tools—OpenAI’s o1 reasoning model and the RAG-powered Vincent AI. The study found statistically significant improvements in quality across several core lawyering tasks like memos, motion practice, and client communications.

📈 This is a real shift from prior studies, which largely showed AI tools improving speed, but not quality. The message for legal teams is clear: when we design our role into the system, we get better outcomes.

🔗 Here’s the post that sparked my interest

I’ve written before about why keeping humans in the loop isn’t just about catching hallucinations—it’s about guiding judgment, shaping outputs, and embedding ethical context. This study shows exactly that in action. When lawyers were paired with the right kind of AI—tools that reason, not just recall, and cite, not just summarize—they delivered higher-quality results.

But the real insight?

🧭 Lawyers aren’t just reviewers. We’re architects of context. The most powerful systems aren’t plug-and-play—they’re built with us in the middle, shaping the questions, interpreting nuance, and setting direction.

AI doesn’t replace legal thinking. It extends it—if we lean into our strengths:

✔️ Asking better questions

✔️ Framing ambiguity

✔️ Exercising judgment

✔️ Communicating with clarity

✔️ Grounding decisions in values

When we do that, AI becomes more than a tool. It becomes a trusted collaborator.

This is the future of product counseling—and of lawyering more broadly. But it won’t happen by accident. It takes intentional design, iterative feedback, and a seat at the table.

Let’s build it that way.

👇 Comment, connect and follow for more commentary on product counseling and emerging technologies. 👇