GitHub Universe ran in San Francisco in early November. The theme was AI across the entire development lifecycle: not just code completion but pull request summarisation, vulnerability detection, and natural language code search.
Copilot Chat in the IDE
GitHub Copilot Chat moved to general availability at Universe. It is a chat interface within VS Code and Visual Studio that answers questions about your codebase, explains code, suggests refactors, and writes test cases. The difference from GitHub Copilot inline suggestions is conversational context: you can ask follow-up questions, reference specific files, and build on previous answers in the same session.
Pull request summaries
GitHub Copilot can now generate a pull request description from the diff. This is a small quality-of-life improvement for individual contributors but a significant workflow change for large teams. PR descriptions that are accurate and complete make code review faster. They also make git history more useful. Generating them from the diff rather than relying on the contributor to write them removes a step that is often done poorly under time pressure.
Copilot Enterprise
GitHub Copilot Enterprise is a new tier that gives Copilot awareness of your private repositories. Instead of training only on public code, Copilot Enterprise indexes your organisation's codebase and uses it as context for suggestions. This is the feature that makes Copilot suggestions aware of your internal libraries, naming conventions, and patterns, rather than just publicly available code.
The AI-native developer workflow
The emerging picture is a developer workflow where: you describe what you want to build in natural language to get scaffolding, you iterate on implementation through Copilot suggestions, you review AI-generated test cases alongside your own, and pull request descriptions are generated from diffs. Each step still requires developer judgment. The AI is compressing the time on the mechanical parts of each step.