GitHub has added a bulk‑start option for its AI Scan service, allowing up to twenty repositories to be scanned in one go. The change is exposed through an organization‑level toggle and the code‑scanning alerts UI, and it lands in public preview on both github.com and GitHub Enterprise Cloud with data residency support. For engineers who already rely on CodeQL, the new bulk capability means faster coverage of default‑branch vulnerabilities across many projects without needing a Copilot license.
How to Initiate Bulk AI Scans
The workflow is two‑fold:
- Enable the organization‑level setting for AI Scan; this unlocks the bulk‑start button on the code‑scanning organization alerts page.
- Select up to twenty repositories and trigger the scan, or start an individual scan from a repository’s own alerts page.
All scans target the default branch of each repository, and the operation is initiated through the web UI rather than a new API endpoint.
Integration with Existing CodeQL Findings
AI Scan complements CodeQL by covering languages that CodeQL does not support and by surfacing additional issues in languages it already analyzes. Results are placed in the same code‑scanning backlog where CodeQL alerts reside, preserving a single view for security teams. This co‑location simplifies triage and prioritisation, as engineers can treat AI‑generated findings alongside traditional static analysis alerts.
Operational and Security Implications
Because the feature is in public preview, organisations should treat it as an early‑access capability: monitor scan duration, backlog growth, and any false‑positive trends. The service is available on both the public GitHub SaaS and Enterprise Cloud, respecting the data residency settings already configured for the account. No additional licensing, such as a GitHub Copilot subscription, is required to use AI Scan.
Related CloudNinjas coverage: security.
What This Means For Practitioners
Teams should evaluate whether to enable the organization‑level setting now or wait for the feature to graduate from preview. If enabled, plan periodic bulk scans to keep the vulnerability backlog current, especially for repositories that lack CodeQL coverage. Keep an eye on the public preview status and any forthcoming guidance on false‑positive handling or performance limits. Integrating AI Scan findings into existing triage pipelines will provide a more complete security picture across your codebase portfolio.
