GitHub’s changelog, dated 5 November 2026, says the Copilot usage metrics API now flags two kinds of gap: users whose editor sends no telemetry, and users running IDE or extension versions below the minimum needed for usage metrics.
What changed
Per-user reports gain two fields. One records whether client telemetry was received that day. The other rates each IDE and Copilot extension as met, unmet or unknown against the minimum supported version. Daily totals also gain counts of users with missing telemetry and users below the minimum version.
GitHub is explicit about the consequence: users without telemetry still count as active, but their reports lack details such as lines of code and feature or model breakdowns. A user can appear in both counts, so they should not be added together. Days before the release have no values.
Why buyers should care
Productivity claims for AI coding tools usually rest on dashboards like this. If part of the team is invisible to the dashboard, any “percentage of code written by AI” figure covers only the developers who are measured. That includes figures a supplier may quote to you.
What to ask a supplier
- What share of the team is actually reporting telemetry?
- Are editors kept on supported versions?
- Are quoted AI gains measured across everyone, or only the visible developers?
So what
Treat AI productivity numbers as a sample until the supplier can show coverage. A supplier who knows their gap is better than one who has never looked. Our AI-assisted development service reports on measured coverage, not headline averages. If you want that in a delivery plan, start a project.