Skip to content
Agentsstrong signalverified

GitHub will turn on unconfigured Copilot features on October 22 unless administrators choose otherwise

GitHub added a global default policy for Copilot Business and Copilot Enterprise on September 24, 2026. From October 22, generally available features that an administrator never configured will follow that default, and the default is Enabled. The policy also covers Copilot code review and MCP servers in Copilot.

By Redakcija WebAiRadarPublished 2 min readwritten by a model
Image: GitHub Changelog

GitHub announced a new setting, Default policy for new features, in the AI Controls section of enterprise and organization settings. You can configure it now, and for 28 days it does not change what your users can reach. On October 22, 2026, it starts deciding the status of every eligible feature left as Unconfigured.

What the policy decides

The policy has three values. Enabled makes current and future eligible features available to users by default. Disabled keeps current eligible features off and requires administrator approval for future ones. Let organizations decide passes the choice to organization owners.

According to GitHub's documentation, the policy itself is enabled by default. An enterprise that does nothing before October 22 will therefore see its unconfigured, generally available features switched on.

What counts as a feature

The policy covers every policy on the enterprise's Features & clients page, plus two more that matter in a security review:

  • the Copilot code review policy on the Agents page;
  • the MCP servers in Copilot policy on the MCP page, which governs whether Copilot can use Model Context Protocol (MCP) servers.

What stays as it is

Explicit decisions are preserved, so a feature an administrator enabled or disabled keeps that setting. Preview features remain opt-in, and a preview someone opted into keeps that choice when it becomes generally available.

Two groups of policies are excluded. One restricts Copilot to data residency or FedRAMP models on GHE.com. The other stores local Copilot CLI and VS Code sessions in the cloud.

Models already follow a similar rule

A separate policy for models, Default availability for released models, is already active. Some models stay disabled whatever it says: pre-release models, open-weight models such as DeepSeek and Kimi K3, and Claude Fable 5 and Claude Fable 5.1. GitHub says those two are not covered by its data retention agreement.

„If you don't take action, unconfigured features will be enabled on October 22.“
GitHub Docs, About default availability of Copilot features and models

Sources

BrandsCopilot

Related

OPENCLAW 2026.10.232,768characters, and past that a key could slip thr
Agentsmedium signal

OpenClaw 2026.10.2 stops long tool output from slipping credentials past redaction

A bug in OpenClaw's secret masking let a credential through whenever it straddled the boundary between two slices of a long text. The fix ships in 2026.10.2 along with repaired updates from 2026.10.1 and a new compaction default for OpenAI and Anthropic API keys.

OpenClawverified

The OpenClaw web interface with a conversation on the left and the Bits & Bolts reference MCP app open beside it, showing a 3D keycap model and an Attach view + note button.
Agentsmedium signal

OpenClaw 2026.10.1 lets agents search installed skills and changes how sessions message each other

OpenClaw released version 2026.10.1 on October 10, 2026, its first stable release since 2026.9.7. Agents can now search the skills installed on a machine and read only the one a task needs, instead of carrying the whole library in the prompt. Opt-in MCP Apps run beside chat, and full backups can go to an external drive or Cloudflare R2. Three changes can break existing setups: session-to-session messaging, the old required argument, and Workboard data stored before July 2026.

OpenClawverified

CLAUDE CODE2.1.294hooks written as instructions now block what t
Agentsmedium signal

Claude Code 2.1.294 fixes prompt hooks that let through the actions they were written to block

Anthropic released Claude Code 2.1.294 on October 8, 2026. It fixes prompt and agent hooks phrased as instructions, which could allow exactly the actions they were meant to stop. Version 2.1.295 followed the same day with an onFailure setting that makes a crashed or timed-out command or HTTP hook block the action. Releases 2.1.292 and 2.1.293 from earlier in the week closed several permission and sandbox gaps.

Claude Code Releases (GitHub)verified