AI Fluency for Lawyers
From occasional prompting to confident, responsible AI use across real legal work.
The aim is not to turn lawyers into AI engineers. It is to make them AI-native: able to decide where AI fits, instruct it well, verify what comes back and use it safely as part of everyday practice.
The guides section shows the style of the teaching.
AI fluency is a legal capability, not a technical specialism.
Lawyers already know how to frame instructions, test ambiguity, interrogate evidence and carry accountability. The training transfers those instincts into a medium that is fast, fluent and capable of being confidently wrong.
Fluency you can demonstrate
Senior lawyers practise candidly, receive direct feedback and show that they can use AI with sound judgment.
Habits, not a prototype
Each session changes how a real task is scoped, instructed or reviewed. The outcome is a repeatable working method.
No engineering curriculum
No code, architecture or automation platform. The goal is confident use, sound judgment and safe delegation.
Six sessions. A different way of working.
Where AI belongs
You identify where AI should assist, where it should not, and who remains accountable for the result.
Instructing with precision
You give a model the context, constraints and output structure needed for reliable legal work.
Reading output critically
You catch omissions, unsupported claims and false confidence, then verify the answer against its sources.
Working AI-first
You use AI across research, drafting, review and intake without handing over legal judgment.
Using it safely
You can explain tenancy, retention, confidentiality and review controls to a client or risk team.
Making fluency repeatable
You leave with a personal working playbook tested on your own matters and ready for daily use.
Executive briefing
Leadership. What to expect, what to fund, what to worry about.
Practice-group workshop
One group, one workflow, a repeatable AI-assisted way of working by the end.
Cohort
Fluency built through weekly practice, office hours and direct feedback. Five to eight people.
In-house programme
Run across practice groups, sequenced to the firm’s own matters.
Hands on keyboards
Participants work on their own legal material, not a prepared demonstration.
Failure taught deliberately
Breaking things on purpose. Recognising a bad output matters more than producing a good one.
Small groups
Five to eight for cohorts. Briefings inform; cohorts turn understanding into confident practice.
A working playbook
Reusable instruction, verification and escalation patterns they can apply after the programme ends.
The objections get honest answers.
“This will replace us.”
It moves the first pass. Lawyers retain review, judgment and accountability.
“It makes things up.”
Ungrounded output is a known failure mode. Source citation, testing and review are built into the workflow.
“We can’t use client data.”
Consumer accounts are the wrong default. Enterprise tenancy, retention and data-use terms are assessed before client data enters the workflow.