Stanford Legal Design Lab

Legal Help Commons

Legal aid organizations, courts, and technology partners across the country are working together to define what legal help AI should do, how to test it, and what quality standards it should meet. The Legal Help Commons is where that shared work lives: reference architectures, conformance standards, evaluation tools, and implementation playbooks that any team can adopt.

Build safe, quality legal help AI together

1

A Shared Agenda of What to Build

The wishlist

Legal aid and court teams are identifying the technical solutions they most need to increase their impact. The Commons is developing these as functional agendas: clear, specific descriptions of what a system should do, what tasks it should handle, and what outcomes it should produce for organizations and the people they serve. This is the kind of clear wishlist that can motivate great, aligned technical development.

See the functional agendas →
2

Testable Standards for Quality and Safety

The goalposts

For each workflow, the Commons is developing conformance standards that define what a technical solution should be able to do in order to meet quality, safety, and sustainability requirements. These standards set the goalposts so builders are aiming at the right targets, public money is being spent wisely, and legal aid teams and their evaluators can tell whether a technology is actually living up to the hype.

See the conformance standards →
3

Blueprints, Guides, Classes, and Case Studies

The how-to

For each workflow, the Commons is developing reference architectures, implementation playbooks, interactive courses, and case studies drawn from real implementations. These are hands-on resources that an internal team or a vendor can follow to borrow from what other teams have already figured out about building effective, safe, and sustainable AI-powered legal help solutions.

Browse the workflows →

What You'll Find in the Commons

Resources to help your team go from "what should we build?" to a deployed, evaluated, and maintained AI tool.

📐

Reference Architectures

Technical blueprints for specific workflows: technology stacks, data flows, integration points, prompt strategies, and decision logic.

📋

Implementation Playbooks

Step-by-step guides covering planning, staffing, procurement, data prep, testing, launch, and maintenance for specific AI workflows.

Evaluation Protocols

Protocols and methods for measuring whether a legal AI tool is accurate, safe, jurisdiction-correct, and useful, before and after deployment.

🔧

Shared Tools and Code

Reusable classifiers, prompt libraries, data pipelines, and integration components that organizations can adopt or adapt.

View the common tools →
📄

RFP Templates and Cost Models

Procurement language, budget frameworks, and staffing models so organizations can plan and fund AI projects realistically.

🎓

Courses and Masterclasses

Interactive classes on how to build, test, and deploy specific legal AI workflows. Designed for mixed teams of legal staff, technologists, and leadership.

Browse courses →
📚

Knowledge Base Preparation

Standards and guides for structuring your legal help content (articles, FAQs, service directories, forms, and legal authorities) so that AI tools can use it reliably. Every AI workflow depends on structured, jurisdiction-specific content.

Read the Knowledge Standard →
🎨

Design Patterns

Tested UI/UX patterns for common justice AI interactions: intake flows, document explanation, multi-step navigation, and more.

Why We Should Build Legal Help Tech Together

When the field builds together, effort compounds instead of scattering, and good work travels from one team to the next.

Build it once

Build a tool once and the whole field can use it, instead of every organization solving the same OCR, classification, and accuracy problems alone.

A shared bar for quality

Shared benchmarks let any team check whether a legal AI tool gives accurate, safe, jurisdiction-correct answers, so the field builds with instruments instead of in the dark.

Lessons that travel

When a team in Illinois learns that a prompt strategy fails for debt collection intake, that lesson reaches the team in Texas facing the same thing.

A level playing field

Shared infrastructure lets a small or rural office stand up capable tools, not only the well-resourced states and organizations.

Legal help workflows the field is building together

Each workflow is an end-to-end scenario where AI can help people with a specific legal problem. Working groups of practitioners, technologists, and researchers are developing the shared agendas, standards, and blueprints for each one. Join a cohort to help shape what gets built.

Standards, courses, tools, and reports

Resources that span multiple workflows. Use these to compare approaches, plan implementations, and learn from work already done.

JusticeBench

JusticeBench is the discovery and evaluation platform for AI in the access to justice field. The Commons builds the workflows and tools. JusticeBench catalogs who is building what, hosts evaluation datasets and benchmarks, and publishes conformance test results.

Explore JusticeBench ↗ Task Taxonomy ↗ Data and Evaluation ↗

Get Involved

The Legal Help Commons is built by working groups of practitioners, technologists, and researchers. Join a working group to help shape the agenda, or sign up to stay connected as new resources, standards, and courses are published.