Our Take
Chen's two-week build exposes how thin the moats are around enterprise legal AI, but without independent benchmarks, 'parity' remains his word against theirs.
Why it matters
Legal departments paying enterprise rates for Harvey and Legora now have a credible alternative to reference in renewal negotiations. Open source availability lets firms address data confidentiality concerns by running models internally.
Do this week
Legal IT teams: Deploy Mike OSS locally this week to test against your current tools before your next Harvey or Legora renewal discussion.
Former BigLaw associate releases Harvey competitor on GitHub
William Chen, a former Latham & Watkins associate, released Mike OSS, an open source legal AI tool that he claims matches the capabilities of Harvey (valued at $11B) and Legora ($5.5B recent raise). Chen announced the launch on LinkedIn, stating he "built their entire web application in two weeks and I'm making it open-source and free for everyone to use."
Mike OSS functions as a legal assistant that creates, edits, and reviews documents using Claude or Gemini models. The tool includes a projects vault for file uploads and colleague sharing, tabular review for bulk document processing, and customizable workflows with preset instructions. The code is available on GitHub, and a demo runs at mikeoss.com.
Chen positions the open source approach as philosophically different from Harvey and Legora, which he calls "black boxes." Mike OSS allows law firms to run and customize legal AI systems locally rather than relying entirely on proprietary platforms.
Enterprise legal AI suddenly has a pricing problem
The release shifts leverage in enterprise legal AI negotiations. James Harrison, former IT Director of Leigh Day, told Legal IT Insider: "Mike doesn't kill Harvey or Legora, but it absolutely changes the negotiation. Once a working open-source alternative is sitting on GitHub, the conversation in renewal meetings moves from 'Is this magic?' to 'What exactly am I paying enterprise prices for?'"
Local hosting addresses a persistent concern in legal AI adoption. Law firms can keep sensitive client data on premises while still accessing AI capabilities. Chen notes that while the public demo site is for testing, firms can download and run the system locally so "your files and data never have to leave your computer."
The broader trend suggests practitioner-built tools may outpace traditional legal software. Shawn Curran, founder of Jylo and creator of open source legal chatbot YCNBot, described this as "vibe coding" where "experts who have practice experience in legal have the power now to write software."
Test locally before your next renewal
Legal departments should evaluate Mike OSS against current tools before upcoming contract renewals. The open source nature allows thorough testing without vendor sales processes or pilot program restrictions.
For firms with strict data governance requirements, local deployment offers a path to AI adoption without external data sharing. However, self-hosting requires internal technical capabilities that not all firms possess.
The tool's rapid development timeline (two weeks, per Chen) and feature set will likely inspire additional open source legal AI projects, potentially creating a more competitive market for enterprise solutions.