GitHub Copilot opt‑out deadline
What happened
GitHub will use interaction data from Free and Pro accounts to train Copilot models starting April 24 — developers must opt out before the deadline or their prompts, code and outputs may be included. This raises IP and privacy flags for proprietary projects and side hustles. (jangwook.net)
Why it matters
Copilot Business and Copilot Enterprise interaction data are explicitly excluded from GitHub’s training program, according to GitHub’s announcement. ) GitHub listed the exact categories it may collect for training, including outputs accepted or modified by users, inputs sent to Copilot (code snippets shown to the model), code context around the cursor, comments and documentation, file names and repository structure, navigation patterns, interactions with Copilot features, and feedback ratings. ) The company clarifies that content from private repositories "at rest" will not be used, but it also says Copilot processes private-repo code while the tool is actively used and that those interaction records could be included in training unless a user changes their preference. ) Opt-out is controlled in GitHub’s Copilot settings on github.com (navigate: profile picture → Copilot settings → Privacy), and GitHub states users who had previously disabled collection will keep that preference. ) ) GitHub’s organizational billing docs list Copilot Business at $19 USD per user per month and Copilot Enterprise at $39 USD per user per month, and those organizational tiers include admin policies and controls that prevent dataset inclusion for organization-managed seats. ) GitHub says it uses filters and de‑identification when training on interaction data, but reporting from security and tech outlets notes GitHub has not published technical details of those safeguards and that the opt‑out default and settings placement have drawn criticism. ) )
Key numbers
- GitHub will use interaction data from Free and Pro accounts to train Copilot models starting April 24 — developers must opt out before the deadline or their prompts, code and outputs may be included.
What happens next
- ) Opt-out is controlled in GitHub’s Copilot settings on github.com (navigate: profile picture → Copilot settings → Privacy), and GitHub states users who had previously disabled collection will keep that preference.
- ) ) GitHub will use interaction data from Free and Pro accounts to train Copilot models starting April 24 — developers must opt out before the deadline or their prompts, code and outputs may be included.
Sources
Quick answers
What happened in GitHub Copilot opt‑out deadline?
GitHub will use interaction data from Free and Pro accounts to train Copilot models starting April 24 — developers must opt out before the deadline or their prompts, code and outputs may be included. This raises IP and privacy flags for proprietary projects and side hustles. (jangwook.net)
Why does GitHub Copilot opt‑out deadline matter?
Copilot Business and Copilot Enterprise interaction data are explicitly excluded from GitHub’s training program, according to GitHub’s announcement. ) GitHub listed the exact categories it may collect for training, including outputs accepted or modified by users, inputs sent to Copilot (code snippets shown to the model), code context around the cursor, comments and documentation, file names and repository structure, navigation patterns, interactions with Copilot features, and feedback ratings. ) The company clarifies that content from private repositories "at rest" will not be used, but it also says Copilot processes private-repo code while the tool is actively used and that those interaction records could be included in training unless a user changes their preference. ) Opt-out is controlled in GitHub’s Copilot settings on github.com (navigate: profile picture → Copilot settings → Privacy), and GitHub states users who had previously disabled collection will keep that preference. ) ) GitHub’s organizational billing docs list Copilot Business at $19 USD per user per month and Copilot Enterprise at $39 USD per user per month, and those organizational tiers include admin policies and controls that prevent dataset inclusion for organization-managed seats. ) GitHub says it uses filters and de‑identification when training on interaction data, but reporting from security and tech outlets notes GitHub has not published technical details of those safeguards and that the opt‑out default and settings placement have drawn criticism. ) )
