jsonscraper
Markets
Back to journal

GitHub lets teams run AI Scan on 20 repositories at once

The new public preview lets organizations start scans in batches, with results appearing in the shared Code Scanning alerts list.

Editorial Policy Report an error

Teams checking the security of multiple repositories can now launch GitHub AI Scan as a batch: on November 16, 2026, GitHub announced a public preview of batch launches for up to 20 repositories. Results appear in the Code Scanning alerts queue alongside CodeQL alerts, so findings can be reviewed in one shared list. GitHub describes the new batch launch in its announcement.

Article: GitHub lets teams run AI Scan on 20 repositories at once
Summary: On November 16, GitHub added batch launches of AI Scan for up to 20 repositories at once. The feature is in public preview on github.com and GitHub Enterprise Cloud with data residency.
Image role: cover — the central idea
Section: 
Visual subject: Several separate software repository folders moving into one shared security review queue, editorial technology illustration showing batch processing and consolidated findings, n
Batch launch lets users start AI Scan for selected repositories; results appear in the shared Code Scanning queue.© jsonscraper · AI-generated

What has changed

Organizations can launch scans for selected repositories simultaneously through organization settings or the organization-level Code Scanning alerts page. For a single repository, launching a scan remains available from its alerts page. The change is therefore about how to start scans across multiple projects, not a new type of result.

Batch launch is in public preview. AI Scan is available on github.com and GitHub Enterprise Cloud with data residency; a separate GitHub Copilot license is not required for this feature. This allows teams in those environments to try launching scans for a selected group of repositories.

What AI Scan covers

According to GitHub, AI Scan looks for vulnerabilities in a repository’s default branch. The company identifies two areas: languages unsupported by CodeQL, and additional issues in languages that CodeQL already analyzes. Findings appear in the Code Scanning queue alongside CodeQL alerts.

For organizations supporting projects in different languages, the ability to start a scan for a selected group through the organization interface and review results in a shared queue is useful. This brings launching scans and reviewing findings together in one process; the security team still needs to assess each alert and decide what action to take.

How to evaluate the new workflow

A team can start with a group of up to 20 repositories, launch AI Scan through organization settings or its alerts page, and then review results in Code Scanning. A single project can still be scanned from its alerts page. A trial run can show how the feature fits into an existing workflow; GitHub’s announcement provides no estimates of detection quality or coverage.

The news is relevant to organizations that need to initiate AI scans across several projects at once, including projects written in languages outside CodeQL’s support. The specific change is batch launch for up to 20 repositories, with results in the existing Code Scanning queue.

People

No people listed for this article yet.

Keep readingOpenAI Opens CSAM Classifier to Approved Organizations: How Image Review Works
Read the next article

Turn what you read into a working integration

Explore jsonscraper's social-data APIs, test requests and build your next workflow.

Explore APIs