BRENNEN

Industries / Finance and business services / Legal and professional services

LegalTech

LegalTech is software for legal work: the tools lawyers use to find the law, draft contracts, sort evidence and run their firms, the systems courts use to take filings and hold hearings, and the services that help people who have a legal problem and no lawyer.

1960

Researchers from the University of Pittsburgh Health Law Center showed lawyers at the American Bar Association's annual meeting in Washington, D.C., a computer searching the full text of Pennsylvania statutes.

michiganlawreview.orgtlp.law.pitt.edu

41

Milestones achieved across 6 domains.

14

To watch for: announced, not yet confirmed outside.

13

Open deliverables. Nobody has shipped them.

In one minute

Where it started

In 1960 a University of Pittsburgh team put Pennsylvania hospital statutes on magnetic tape and showed lawyers a computer searching every word of them. Courts began taking filings over the internet in 1997, in Singapore and in a US federal court in Ohio. Help for people without lawyers began on paper, with self-help law books from Nolo Press in 1971, and moved to online interviews that fill in court forms.

Where it stands

Electronic filing is routine in many countries: in 2023 only 0.4% of new cases in Brazil's courts were filed on paper, by the national justice council's count. AI tools spread fast, and 30% of lawyers in the American Bar Association's 2024 survey said they used them, up from 11% a year earlier, yet in an independent run of a test of legal AI agents the best model fully finished 14.2% of 120 tasks. Help for people without lawyers lags behind: in a 2022 US study, low-income Americans got no legal help, or not enough, for 92% of their serious civil legal problems.

What is open

The field lacks shared test data more than it lacks tools. Most open jobs below are public test sets and public records turned into data: dated versions of state statutes, a chat collection for discovery review, a table of court filing fees, and tests of contract deadlines, conflict checks and form finding. Each can be started this month with public records or invented examples and current AI models, and none gives legal advice to a real person.

Domains

AchievedWatch forOpen

Advances to build with

2 research notes

Open in legaltech

3 of 13 open deliverables

Open · Research

Dated state statutes as data

For one US state, an open dataset of every section of the state's code as it read in each past annual edition, with the date each version took effect, so that anyone can answer "what did section X say on date Y". Publish it as a git repository with one commit per edition, and a simple lookup page. It records text and gives no advice.

Built with what exists

state legislature archives and the Internet Archive's Wayback Machine for past editions; Docling or pdfplumber, open-source tools that pull text out of PDFs and scans; a language model such as Claude, Gemini or Meta's Muse Ember to split each edition into sections, with a person checking (see the Note "The jobs agents repeat"); git to record each change.

Credit · unclaimed

Open · Contract tools

A notice deadline test

A test in which an AI model reads a renewal or termination clause and a start date, works out the exact deadlines, such as the last day to give notice that a contract will not renew, and writes them as calendar entries (.ics files). Score exact date matches.

Built with what exists

EDGAR full-text search, the SEC's free search of company filings, where material contracts appear as exhibits; Stanford's Material Contracts Corpus; Python date and holiday libraries and an iCalendar library; Claude, Gemini and GPT models.

Credit · unclaimed

Open · Discovery

A modern chat test collection

A free public test collection of invented workplace chat, so that review tools can be scored on today's data instead of old email: Slack- and Teams-style channels and direct messages with threads, edits, deletions, emoji reactions and file links, plus at least 20 mock document requests and relevance labels made by people.

Built with what exists

OrgForge (March 2026), an open-source tool that simulates a company and writes its Slack messages, email and meeting transcripts while keeping a log of what really happened, but has no document requests or relevance labels; Relativity Short Message Format (RSMF), a documented format that review tools use for chat; Claude, Gemini, GPT or open-weight models to write message text; two or more people to label relevance.

Credit · unclaimed

Where it sits