License GitHub data to AI: repos, history and checks
The short answer
To license GitHub data to AI, know what Troveo's guide says labs want: history, pull requests, code review and issues. Troveo's guide cites an unnamed platform's $5,000 baseline per repo. Troveo says buyers check that every contributor assigned work to you. Counsel clears the license and secrets side.
Code is the asset startups license first when they shut down. micro1's data partnership asks about it too: David Revland, who works on human data at micro1, named GitHub in an undated recorded forum talk.
What does Troveo say labs want from a GitHub repository?
The whole development story of a GitHub repository, per Troveo's guide: full commit history, pull requests and code review, and issues and tickets linked to code. They show how decisions got made.
SimpleClosure's Asset Hub, launched 16 Apr 2026, lists source code as its first product and says it licenses codebases to vetted buyers.
The Spirit Airlines archive carried about 30 million lines of code in the archive behind Google's $10M winning bid, pending court approval. My Spirit Airlines page has the status.
What does a repository earn?
Troveo's guide cites $5,000 per repository as an unnamed platform's baseline, adjusted after grading. Troveo does not name the platform.
SimpleClosure said its nearly 100 deals paid $10,000 to $100,000 per company, which covers code and workspace data together (Gizmodo).
Troveo says in another guide that there is no public price list. My deals tracker logs public prices as they appear.
Does micro1 ask about GitHub?
micro1's David Revland named GitHub in an undated recorded forum talk. It is one of the tools the data manifest form asks about.
Who owns the code?
GitHub's terms say you own your content and that GitHub treats private repository contents as confidential to you. The terms bar using the API to download content in order to sell GitHub users' personal information.
Ownership on GitHub's side is the easy part. Troveo says buyers verify that every contributor, including contractors and agencies, assigns work to the company.
Frankfurt Kurnit Klein & Selz lists licensed code, open source and vendor code among items needing careful review.
What about a startup's code?
A shutting-down startup holds the same repositories, plus the stockholder and creditor questions on my shutdown page. The legal side of code, including the Ninth Circuit's 16 Sep 2026 ruling in Doe v. GitHub and export controls, is on can I license source code to AI.
Talk to a commercial lawyer in your state before you sign.
micro1 may pay us a referral fee if a company we introduce goes ahead with it. How we are paid
Apply to micro1's data partnership (paid link)micro1's own data-partnerships pagePaid link. How we are paid
Vaultfish helps you run the process. We are not lawyers and this is not legal advice. If you need counsel who knows this area, we can point you to one. Ask for a referral
