The legal profession is, right now, deciding what counts as AI competency. You can feel it in the rooms: bar committees, the PLI Innovation Council and the Professional Development Consortium building the AI-Ready Lawyer framework, law schools rewriting modules, firms drafting the skills their associates will be measured on. It is overdue, and it is the right conversation to be having.
But almost all of it points at one target: the individual lawyer. How they verify outputs. What they should know. Which skills count now. That half matters. It is also only half.
Competent people, indefensible system
Take the two lawyers recently hit with a combined $110,000 sanction in Oregon, one of the largest US penalties yet for AI-fabricated citations. You can read it as a simple competence failure, a lawyer who did not check. The more useful reading is structural: nothing in the workflow forced a verification step, so twenty-three fabricated citations and eight invented quotations travelled all the way to a court filing. Better-trained people help. A workflow that makes verification unavoidable is what actually stops this.
That is what the individual-skills lens misses. You can train every lawyer in the building and still produce work you cannot defend, if the system around them treats verification as an optional last step.
The reframe
The most important idea coming out of the AI-Ready Lawyer work is this: verification cannot be a gate at the end. By the time the work reaches a final review, the reasoning has already happened. Verification has to live inside the workflow, at every point where AI touches the work.
And the moment it lives in the workflow, competency stops being purely individual. It becomes a property of the operating model. So being AI-ready is really two questions asked at once. Is the lawyer ready? And is the operating model around them ready? The first is getting a framework. The second is where defensible work actually comes from.
So which problem do you have?
We come back to one distinction constantly: capability is not maturity. Capability is whether your people can use AI. Maturity is whether you can govern it, evidence it, and defend it. Most legal functions now have capability. Very few have maturity, and the gap between the two has a name. We call it Exposed: high capability, low defensibility. It is the most common place to stand, and the most dangerous, because it feels like progress.
An AI-ready lawyer inside an exposed function is still doing exposed work. The individual is ready. The system is not.
Before you adopt anyone's competency checklist, it is worth knowing which of the two problems you have. A trained lawyer in an ungoverned system needs a different fix than an under-trained lawyer in a sound one. The first step is to locate your function honestly. The Free Baseline does this in five minutes, at no cost.
The deeper read: Diagnostic Pro
If the baseline surfaces a gap you want measured properly, Diagnostic Pro is the four-lens instrument: a fifty-question assessment scored across Adoption, Sophistication, Defensibility, and Autonomy, then read personally by our founder. It returns your Maturity Band, your lens profile, and a prioritised path toward defensible AI use.
We are opening it now to a founding cohort. Founding members receive the founder-reviewed Diagnostic Pro Report, founding-member terms, and a direct line into how the instrument develops. The first cohort is limited, and this is the moment to take a place in it.
The profession is finally building real standards for the AI-ready lawyer. The firms that win the next two years will be the ones that build the AI-ready operating model to put them in.
Nishant Bhaskar · Founder, Advanta
Share this issue