What happened
Microsoft has issued security patches addressing a critical flaw in Azure AI Foundry, designated as CVE-2026-85889. This vulnerability results from missing authentication for a critical function within the Azure AI Foundry platform. Exploiting this flaw could allow an unauthorised attacker to escalate privileges over a network, potentially gaining elevated access to affected systems. The vulnerability’s CVSS score has been rated at 10.0, indicating maximum severity and an ease of exploitation that poses significant risk if left unmitigated. Microsoft confirmed that there is no required action from affected organisations to implement these patches, as the updates are automatically applied.
Who is affected
Organisations running Microsoft Azure AI Foundry are impacted by this security flaw. The vulnerability affects deployments where the platform’s critical functions are exposed without sufficient authentication controls. As no particular patch version or environment is specified beyond the affected component, all instances of Azure AI Foundry are considered vulnerable until the available updates are applied.
Recommended actions
- Ensure that system patches related to Azure AI Foundry have been applied as provided by Microsoft, since no manual customer action is required, and updates are automatically distributed.
- Review and monitor network traffic for any suspicious activity indicative of exploitation attempts, especially around functions that do not enforce authentication.
- Implement standard network security measures, including segmenting affected services and maintaining robust perimeter protections, to reduce exposure to unauthorised access.
- Maintain an activity monitoring and alerting framework capable of identifying abnormal privilege escalations or anomalous access patterns.
Indicators of compromise
Specific indicators of compromise have not been provided at this time.
