About JusticeBench
JusticeBench is an initiative by the Stanford Legal Design Lab to help researchers, technologists, and public service leaders evaluate and build better AI systems for civil legal help. The Lab also leads the Legal Help Commons, and JusticeBench is the discovery layer of the Legal Help Commons.
JusticeBench shows you what’s possible. The Legal Help Commons shows you how to do it, with implementation playbooks, reference architectures, and working groups you can join today.
This site hosts projects, benchmarks, datasets, and community-contributed tools that aim to ensure AI can responsibly support access to justice, especially in domains like eviction, debt, custody, and government benefits.
It was created by a team at Stanford Legal Design Lab, supported by partners across the U.S. justice system, and is intended to grow as an open platform for knowledge sharing and testing.
JusticeBench’s goals within the Legal Help Commons:
- Make AI projects in justice more transparent and measurable
- Help legal aid and court teams find and adapt existing AI projects
- Create shared benchmarks to improve quality and safety
- Foster responsible, inclusive innovation for legal help
- Connect teams to the implementation playbooks, reference architectures, and working groups of the Legal Help Commons