product updateGitHub

GitHub will train Copilot models on user interaction data starting April 2026

TL;DR

GitHub will use Copilot interaction data from Free, Pro, and Pro+ plan users to train AI models starting April 24, 2026, unless users actively opt out. The policy does not affect Copilot Business and Enterprise customers. Data shared will include prompts, outputs, code snippets, filenames, and repository structures.

2 min read
0

GitHub Will Train Copilot Models on User Interaction Data Starting April 2026

GitHub announced a significant change to its Copilot data policy effective April 24, 2026. Starting that date, interaction data from users on Free, Pro, and Pro+ plans will be used to train AI models unless users explicitly opt out.

What Data Will Be Collected

The data collection will include:

  • User prompts and model outputs
  • Code snippets
  • Filenames
  • Repository structures
  • User feedback on model suggestions

Users who have previously opted out of data collection will retain their existing settings and will not be automatically enrolled.

Scope and Limitations

Copilot Business and Enterprise customers are exempt from this policy change. GitHub clarified that collected data can be shared with Microsoft but will not be shared with third-party AI model providers.

GitHub Chief Product Officer Mario Rodriguez stated that real-world usage data improves model quality. Internal testing with data from Microsoft employees already demonstrated higher acceptance rates for code suggestions, suggesting the approach yields measurable improvements.

Opt-Out Process

Users who wish to prevent their data from being used for training can opt out through Copilot settings under the "Privacy" section. GitHub indicated that more details are available on the GitHub blog.

What This Means

This policy represents a shift toward extracting training value from Copilot's large user base of developers. The limitation to Free, Pro, and Pro+ plans—while exempting Enterprise customers—suggests GitHub is balancing competitive advantage with enterprise customer expectations around data usage. The explicit opt-out structure (rather than opt-in) will likely result in significantly higher participation rates unless developers proactively change settings. The exclusion of third-party providers indicates Microsoft intends to keep this training data advantage internal rather than licensing it broadly, positioning GitHub Copilot improvements as a Microsoft-exclusive benefit.

Related Articles

product update

GitHub Adds Agent App Activity Data to Copilot Usage Metrics API

GitHub has updated its Copilot usage metrics API to include activity data from agent apps such as Claude and Codex operating within GitHub workflows. The change gives organizations visibility into third-party agent usage alongside native Copilot metrics.

product update

Amazon, Cursor, Microsoft, OpenAI, and Vercel Launch Agent Plugins, a Shared Packaging Standard for AI Agent Extensions

Amazon, Cursor, Microsoft, OpenAI, and Vercel have released Agent Plugins, an open standard defining a single package format for AI agent extensions. Version 1.0.0 covers Agent Skills and MCP servers, but leaves marketplaces, permissions, and runtime out of scope.

product update

GitHub Publishes Guide to Slash Commands in the Copilot App

GitHub has published a guide covering slash commands available in the GitHub Copilot app, designed to extend Copilot beyond simple chat into planning, team collaboration, task automation, and workflow customization. The guide targets developers looking to get more structured, repeatable value out of Copilot's interface.

product update

Anthropic Makes Auto Mode Default in Claude Code for Pro, Max, and Team Plans Starting August 14

Anthropic will make auto mode the default setting for new Claude Code sessions on Pro, Max, and Team plans starting August 14, 2026. The company cites a 1,053-person study showing auto mode blocked 89% of harmful actions compared to 13.6% for human reviewers, plus a third-party test claiming zero successful prompt injections out of 720 attempts.

Comments

Loading...