Can a Software Company Sell or License Its Codebase to AI?
AITrainingDataBroker.com
Can a software
company sell
its codebase?
The history of how your code was built may be worth more than the code.
Jimmy Lin
Can a software company sell its codebase?
Hello, everyone. If your company writes software, you are sitting on something most owners never think about: the full history of how your code was built.
In this video, I will show you why AI companies may want that history, what 2 studies say about it, and the checks to run before anyone sees a line.
The video
Full script
Hello, everyone. If your company writes software, you are sitting on something most owners never think about: the full history of how your code was built.
In this video, I will show you why AI companies may want that history, what 2 studies say about it, and the checks to run before anyone sees a line.
Picture the last bug your team chased for 3 weeks. The first guess failed, the second guess made it worse, a reviewer asked a sharp question, and then the fix finally landed.
All of it is saved, commit by commit, in your repositories, which is the place where your code and its whole history live.
Most owners see that as an old archive, but to an AI company it can look like a recording of how a good engineer thinks.
I'm Jimmy. I have worked in data at Fortune 500 companies, as a data analyst, a data engineer and a software engineer, and I am trained in data science. I am also a Canadian military veteran, where I specialized in data analysis and reporting.
Are you selling your data? We are your broker and technical advisor. We get you the best price, and we guide your team through the technical side, so your data stays private.
Are you buying data for an AI lab? We find the right data for you and see it through to delivery, so you can focus on advancing AI research.
Start with the difference between a snapshot and a story.
The code you ship today is a snapshot, and an AI company can already read a huge amount of public code.
What it can not read is your private history: why each change was made, what was tried first, and what a reviewer pushed back on.
That is the hard part to find, and your team already has it.
So what exactly is in that history?
A commit is a saved change with a short note about why. A pull request is a proposed change that a teammate reviews, and the review comments are often the best part, because that is where a senior engineer explains what is wrong.
An issue is the bug report or request that started the whole thing.
Put those 3 together and you get the problem, the discussion and the fix in one trail.
Is that just a nice idea? Researchers say there is something to it.
In a paper called OctoPack, presented at ICLR, a major AI research conference, in 2024, a team led by Niklas Muennighoff used about 4 terabytes of Git commits, each one pairing a code change with the note a person wrote about it.
They used those pairs to teach code models to follow instructions, and the models did better than with the other instruction sets they tried.
In plain words, a commit is a ready-made lesson: here is the problem, and here is the fix.
A second study looks at the same idea from the testing side.
SWE-bench, by Jimenez and colleagues, also at ICLR in 2024, built 2,294 tasks from real issues and the pull requests that fixed them, across 12 popular open-source Python projects.
When it came out, the best AI solved fewer than 2% of them.
So real issues with real fixes are both how AI learns software work and how it gets graded, which is why a private history of them is scarce, and useful.
Now for the question every owner asks: is it even mine to license?
For code written by your employees, it usually is, as long as their agreements assign the work to the company.
For contractors, check the same thing, because it is not automatic.
Code built for a client often belongs to the client, and open-source code comes with its own license, which you have to follow.
This is not legal advice, so ask your own lawyer, and I work through the rest with you.
Then there is the risk that surprises engineers.
Old commits can hold passwords and keys that someone saved by mistake and later removed from the current code, but not from the history.
GitGuardian, a company that scans code for exactly this, reported in 2025 that 35% of the private repositories in its scans held at least one plaintext secret.
So before any AI company sees anything, we find and remove those secrets, and your engineers change the keys anyway, which we guide your team through.
Put it together, and it looks like a short readout.
Is the code yours, are the contractor agreements in place, is any client code mixed in, which open-source licenses apply, and what is hiding in the history?
Each line is either fine or something to fix first, and most companies have a few of the second kind, which is normal.
And what do you give up? Not the code.
A license means the AI company gets limited rights to learn from the history, while you keep your code and keep shipping your product.
The contract sets the limits on how it is used, and who pays if something goes wrong.
I work through the terms with you, and you make the final call.
So, is your code history worth more than you thought? Now you know what to check first.
If you run a company with 30 or more people, the data you already have could become a new revenue stream. You can find out if yours qualifies, for free, at AITrainingDataBroker.com.
If you have further questions, leave them in the comments, and I will answer them in the next video.
Thank you for watching, and I will see you next time.
More slides
AITrainingDataBroker.com
Which systems
make money
from AI?
Email, files, chat, your CRM and accounting: which ones AI companies want.
Jimmy Lin
AITrainingDataBroker.com
What sets
the price
of your data?
There is no price list. There are 4 questions every AI company asks first.
Jimmy Lin
AITrainingDataBroker.com
Can a real estate agency
sell or license
its data to AI?
For real estate agents, teams and brokerages: what AI still needs from you
Jimmy Lin
