Today, April 24th, the deadline has passed. If you’re using GitHub Copilot Free, Pro, or Pro+ and you haven’t touched a specific setting in your profile, your interaction data is already being used to train AI models — by default, without your explicit authorization.
This isn’t a rumor or a panic thread on Reddit. It’s GitHub’s official policy, announced on March 26th by Chief Product Officer Mario Rodriguez: as of today, everything you do with Copilot on an individual plan becomes training data. The prompts you write, the suggestions you accept or reject, code snippets in your active sessions, file names and folder structure visible while you work — all of it.
What data is collected exactly
GitHub clarifies that it doesn’t directly train on code from your private repositories “at rest”. But that distinction matters less than it seems. When you actively work in a private repo with Copilot open, the interaction data from that session — including code context visible on your screen — is part of it. A single developer using Copilot in a private codebase exposes the interaction patterns of that entire project.
The scope includes: code inputs and outputs, accepted and rejected suggestions, file context during active sessions, navigation patterns, chat messages, and feedback signals (thumbs up/down). It’s a pretty broad surface.
Who’s affected and who isn’t
Copilot Business and Enterprise users aren’t affected — their contracts explicitly exclude it. Students and educators accessing Copilot through academic programs are also exempt.
If you’re on Free, Pro, or Pro+, you’re in the pool unless you’ve opted out. And if you previously chose not to participate in GitHub’s general product improvement data collection, that preference transferred automatically — you don’t need to do anything. But if you’re unsure, check anyway.
How to opt out (takes 30 seconds)
Go to: github.com/settings/copilot/features
Under the Privacy section, turn off “Allow GitHub to use my data for AI model training”.
That’s it. GitHub confirmed that opting out doesn’t affect access to any Copilot functionality.
Why this matters beyond your personal settings
The governance angle is what most developers are overlooking. Even if you opt out, what happens when a teammate — working on the same private repository — doesn’t?
For teams handling regulated data, proprietary algorithms, or client code, this isn’t a question of personal preferences: it’s a supply chain risk question. The right answer for most engineering organizations is a team policy, not an individual toggle.
The underlying signal: GitHub is making this move because high-quality interaction data — what real developers accept versus what they reject — is genuinely scarce and valuable. Public repositories as a training source are practically exhausted. Real coding sessions are the next frontier. That’s the business logic behind why the option comes enabled by default.
Whether that trade-off is acceptable is a decision every organization should make explicitly — not by omission.
Check your settings. Then check your team’s.
Direct link: github.com/settings/copilot/features