Azure Pipelines allows remote code execution by unauthorized attackers
Azure Pipelines is Microsoft's service that runs software build and deployment jobs. A flaw lets an outside attacker, with no login, send specially crafted input over the network and make the service run commands of their choice.
| Severity | CriticalCVSS 9.1 of 10 |
|---|---|
| Being exploited? | UnclearNo reports either way yet |
| Fix available? | UnclearNo fixed version published in the sources yet |
| How sure is this? | Official, 3 sourcesBacked by an official source |
What happened
Improper neutralization of special elements in Azure Pipelines leads to command injection, letting an unauthorized attacker execute code remotely over a network.
Are you affected?
You are affected if you use Azure Pipelines in an affected version; the material does not specify exact versions or configurations.
Azure Pipelines (specific versions unknown)
What to do
- Check the Microsoft MSRC advisory for CVE-2026-105014 for the official fix or mitigation for Azure Pipelines
- Apply the patch or mitigation Microsoft specifies as soon as it is available to you
- Review Azure Pipelines access logs and restrict network exposure where possible
Timeline
- 22 Oct 2026, 14:00 UTCPublicly disclosed
- 22 Oct 2026, 14:00 UTCReported by Microsoft MSRC · Microsoft MSRC
- 22 Oct 2026, 22:17 UTCReported by NVD · NVD
- 25 Oct 2026, 17:05 UTCGitHub Advisory reported on it · GitHub Advisory
Sources 3 reports, oldest first
Official vendor, government CERT or advisory database
- CVE-2026-105014 Azure Pipelines Remote Code Execution VulnerabilityMicrosoft MSRC · 22 Oct 2026, 14:00 UTC
- NVD22 Oct 2026, 22:17 UTC
- Improper neutralization of special elements used in a command ('command injection') in Azure...GitHub Advisory · 23 Oct 2026, 00:33 UTC
Technical details
- ID
- CVE-2026-105014 · GHSA-vg36-vcf8-m9cg
- CVSS
- 9.1 / 10
- EPSS
- 0.7% chance of exploitation within 30 days
- Disclosed
- 22 Oct 2026, 14:00 UTC
- On Zero Radar
- before the radar started
- Last change
- 25 Oct 2026, 17:05 UTC
The plain-language parts and the attack diagram were written by an AI model (anthropic/claude-haiku-5.5) from the sources above. Facts like versions, exploitation and dates come from the sources themselves. How Zero Radar works