Skip to content
Legal AI Laboratory

About

About Legal AI Laboratory

An independent lab for building, testing, and documenting AI systems for legal work.

Build → Test → Document → Share

01 · Why It Exists

Legal AI is moving quickly. Useful legal AI is harder.

Find out what AI can actually do in legal work — by building systems, testing them on legal tasks, understanding where they work and fail, and figuring out how attorneys can work with them.

  • What can AI actually do in legal work?
  • Which legal tasks can AI perform well — and which remain difficult?
  • How should AI systems and workflows be structured to make legal work more reliable?
  • Where does AI succeed, where does it fail, and why?
  • Where does attorney judgment and supervision remain essential?
  • What can we learn by building, testing, and documenting these systems in practice?

02 · What We Do

What We Do

Four activities, repeated as a loop.

01

Build

Build AI systems, pipelines, AI agents, and legal workflows.

02

Test

Test those systems using structured legal tasks.

03

Document

Document experiments, notes, benchmarks, and lessons.

04

Share

Share tutorials, resources, and teaching materials.

↺ Then build again.

03 · Independence

An independent project.

Legal AI Laboratory is an independent project. It is not affiliated with any law firm, university, legal technology company, or other institution.

Legal AI Laboratory does not represent clients or provide legal advice through this website. The content presented here reflects research, experimentation, and observations about AI and legal work. It is not legal advice and should not be treated as an attorney-reviewed legal work product.

Project Philosophy

  • Show the work

    Show what was actually built, tested, and observed.

  • Separate evidence from conclusion

    Treat observations as evidence, not as proof of a broader claim.

  • Make uncertainty visible

    When the answer is unknown, say so.

  • Document failure

    Failures, errors, and unexpected results are part of what we learn.

  • Prefer evidence over hype

    Make claims based on what we observe, not what we hope.

  • Keep humans accountable

    AI may assist legal work, but attorney responsibility remains clear.

04 · The People

The People

The Laboratory is built at the intersection of law, engineering, and artificial intelligence.

Portrait of Dingkun Ren

Dingkun Ren, Ph.D.

Patent Attorney · Admitted in California and the District of Columbia

Dingkun is an intellectual property attorney and a former hardware engineer in Silicon Valley with a Ph.D. in Electrical Engineering. His work spans legal practice, complex technologies, and artificial intelligence.

Areas of Work

  • Intellectual Property Law
  • Artificial Intelligence
  • High-Tech
Portrait of Xu Li

Xu Li, Ph.D.

Patent Attorney · Admitted in California and the District of Columbia

Xu is an intellectual property attorney and a former hardware engineer in Silicon Valley with a Ph.D. in Mechanical Engineering. Her work spans legal practice, complex technologies, and artificial intelligence.

Areas of Work

  • Intellectual Property Law
  • Artificial Intelligence
  • High-Tech