How should a law firm train its lawyers to use Claude?
By Michael Showalter ·
Short answer
Train lawyers on a repeatable workflow. The firms that get real productivity gains teach a repeatable method for directing the model on actual matters, and then encode the firm’s own standards so every lawyer starts from the same baseline.
Most AI training for lawyers is awareness training: what the tools are, what the risks are, and a few sample prompts. That is useful background, but it rarely changes how anyone works on Monday. The training that changes behavior is built around the work lawyers already do.
An effective program covers four things:
- Setting up the matter. Give the model the record, the controlling authority, and the firm’s drafting standards before it starts drafting.
- Directing iteratively. Research, draft, and revise in rounds, the way a partner directs an associate.
- Building in self-checks. Have the model test its own work and flag what it is unsure of.
- Lawyer sign-off. A lawyer is responsible for the final work product.
The second half of training is firm-wide: turning the firm’s procedures, templates, and style rules into reusable instructions (“skills”) so that every lawyer’s sessions start from the firm’s standards.
Who teaches it matters. Lawyers learn fastest from practitioners who use these methods on their own matters and can show what good use looks like on real legal work.
Showalter AI Consulting runs workshops on this method, from a 60–90 minute keynote to a full-day program with practice-group breakouts.
Showalter AI Consulting is not a law firm and does not provide legal services or legal advice. Our services do not create an attorney-client relationship.