How to Stop GitHub From Using Your Data to Train Copilot’s AI Models

GitHub is arguably the world’s most popular code storage and management platform. I use it on a daily basis myself, and its collaboration tools ensure that I can work with other developers on the same code without fear of conflict or mangling each other’s changes. It’s pretty much an indispensable tool, and I can’t really imagine my job without it.

Since 2021, GitHub — which is owned by Microsoft — has been steadily incorporating AI features, and specifically, their GitHub Copilot coding assistant, which can be used to generate code with simple user prompts. Like all AI tools, though, Copilot needs to be trained on something in order to be useful. Which is why GitHub has announced that on April 24, they will begin training Copilot on user data and activity:

From April 24 onward, interaction data — specifically inputs, outputs, code snippets, and associated context — from Copilot Free, Pro, and Pro+ users will be used to train and improve our AI models unless they opt out. Copilot Business and Copilot Enterprise users are not affected by this update.

In the related FAQ, GitHub clarifies that Copilot will only be trained on developers’ Copilot interactions, and not their code “at rest.” In other words, if you only use GitHub to archive your code in private repositories, then it won’t be touched. Regardless, it’s easy to opt out of this and keep your work to yourself.

1. Log in to your GitHub account.
2. Go to the “GitHub Copilot” settings.
3. Scroll down to “Allow GitHub to use my data for AI model training” and set it to “Disabled.”

When you select “Disabled,” GitHub displays a little warning that you’ll no longer have access to this particular feature. Now, I fail to see how having my work gobbled up by an AI model for the benefit of a multi-billion-dollar company is a “feature,” but maybe that’s just me.