News date: August 18, 2026. GitHub has added enterprise managed settings to GitHub Copilot for JetBrains. Enterprise owners can now apply centrally defined controls for Copilot plugins, MCP server access, OpenTelemetry, and agent permission modes inside supported JetBrains IDEs.

What GitHub announced

  • Plugin governance: administrators can require or disable specific plugins, add approved marketplaces, and restrict installation to known marketplaces.
  • MCP controls: allow and deny lists can limit which Model Context Protocol servers Copilot connects to.
  • Managed telemetry: enterprises can set the OpenTelemetry collector, protocol, service name, resource attributes, and content-capture policy. Managed values override developer settings.
  • Approval enforcement: administrators can prevent the JetBrains agent from using Bypass Approvals or Autopilot modes.

Why this matters for QA engineers

AI-assisted testing increasingly depends on external tools: test-management plugins, browser or database MCP servers, terminal commands, and telemetry pipelines. A locally configured agent can create inconsistent risk across laptops. JetBrains support lets an enterprise apply the same baseline controls to QA engineers who work in IntelliJ IDEA, PyCharm, WebStorm, and related environments.

The update also gives QA teams a testable policy layer. Instead of assuming a restriction works, testers can verify that blocked MCP endpoints stay blocked, unapproved marketplaces cannot supply plugins, telemetry reaches only the approved collector, and agent commands still stop for human approval.

A focused QA validation plan

  1. Pilot on a small device group: GitHub’s deployment guidance explicitly recommends a limited rollout before broad distribution.
  2. Test allow and deny precedence: configure one permitted MCP server and one denied server, then verify both expected connection paths.
  3. Try an unapproved plugin source: confirm restricted marketplaces fail closed and produce a useful message.
  4. Validate approval gates: ask the agent to run a harmless command, edit a disposable file, and fetch a test URL; confirm bypass modes remain unavailable.
  5. Check telemetry routing: verify the configured service name and resource attributes at the approved collector, while confirming the content-capture policy is respected.
  6. Measure policy propagation: server-managed settings normally apply within about an hour; restarting the client or signing in again should trigger an immediate refresh.

Deployment and precedence

GitHub documents server-managed, MDM-managed, and file-based deployment options. In the published precedence order, MDM settings take priority over server-managed settings, followed by file-based and user-level settings. Teams should include conflicting-value tests in rollout validation so the effective policy is predictable on every managed workstation.

Bottom line

Copilot’s JetBrains governance update closes an important consistency gap for enterprises with mixed IDE fleets. For QA leaders, the practical next step is a controlled policy test matrix covering plugins, MCP access, telemetry, approval behavior, and refresh timing before enabling agent workflows widely.

Sources: GitHub Changelog: Enterprise managed settings in GitHub Copilot for JetBrains (August 18, 2026); GitHub Docs: Configuring enterprise-managed settings; GitHub Docs: Enterprise managed settings reference.