Tools / GitHub Copilot

GitHub · Copilot Free, Pro, Pro+ and Max

How do I stop GitHub Copilot from training on my data?

On GitHub, open your profile picture → Copilot settings and set “Allow GitHub to use my data for AI model training” to Disabled. Since April 24, 2026, interactions (inputs, outputs, code snippets, context) on these plans may be used for training unless disabled, and Business and Enterprise plans aren’t used for training and don’t show the setting.

Last checked: Last verified: Source: official GitHub documentation

What are the steps?

  1. On any GitHub page, click your profile picture.
  2. Click Copilot settings.
  3. Set “Allow GitHub to use my data for AI model training” to Disabled.

Items marked Being verified aren’t clearly stated in the official documentation yet. We’re confirming them before adding detail. Spotted a change? Tell us.

Cover of the Small-Business AI Privacy Checklist.

$24 · PDF

Small-Business AI Privacy Checklist

This page covers one person’s GitHub Copilot account. The Small-Business AI Privacy Checklist covers training controls, account types, and a team policy template.

Get the team checklist ($24)

For the household: Family Privacy Setup Kit ($9)

Our PDF, separate from the steps above. The guides on this site stay free.

What does turning it off actually do?

  • Since April 24, 2026, interactions (inputs, outputs, code snippets, context) on these plans may be used for training unless disabled. Business/Enterprise data not used; setting hidden on those plans.

How long is my GitHub Copilot data kept, and how do I delete it?

Being verified — see GitHub Privacy Statement linked from the above page.

Claude Fable 5.2 is available to Copilot Pro+, Max, Business, and Enterprise, and GitHub says the rollout is gradual. Using Claude Fable 5, Fable 5.1, or Fable 5.2 means Anthropic retains prompts and outputs by default to operate safety classifiers, outside GitHub's usual zero-data-retention agreement with Anthropic unless an eligible enterprise is approved for a time-bound exemption through the end of 2026. GitHub says that retained data is not used to train Anthropic's models. Copilot Business and Enterprise admins must enable the Claude Fable 5.2 policy in Copilot settings; the policy is off by default. Source: https://docs.github.com/en/copilot/reference/ai-models/model-hosting , https://github.blog/changelog/2026-11-09-claude-fable-5-2-in-github-copilot/

Is Copilot data used for training unless I disable it?

Since April 24, 2026, interactions (inputs, outputs, code snippets, context) on these plans may be used for training unless disabled.

Which official pages were checked?

  1. docs.github.com/en/copilot/how-tos/manage-your-account/manage-policies
  2. docs.github.com/en/copilot/reference/ai-models/model-hosting
  3. github.blog/changelog/2026-11-09-claude-fable-5-2-in-github-copilot/

Checked 2026-11-11. Not affiliated with GitHub. Not legal advice.

This setting has changed before and may again. Email me when it changes.