Free practical cybersecurity guidance for organisations without a security team.
hello@actionsoncyber.com
← Back to Vulnerability Briefs

Critical Microsoft Azure AI Security Flaw Allows Unauthorized Access

A critical security flaw has been found in Microsoft’s Azure AI Foundry that could let attackers gain higher access rights without permission. This matters because it could allow someone to take control over parts of your cloud services, potentially leading to data breaches or disruption.

20 September 2026

Reference: CVE-2026-85889

1. What is being reported?

The report describes a serious security weakness in Azure AI Foundry where a key function does not require proper authentication. This means an attacker on the network could exploit this to increase their access privileges without needing to log in properly.

2. What this means in plain English

For small organisations using Microsoft Azure AI services, this vulnerability could allow hackers to access sensitive information or control systems they shouldn’t. This could lead to data loss, service interruptions, or other security problems.

3. Could this affect a small business?

Small businesses or charities using Azure AI Foundry or related Microsoft cloud AI services could be affected. Those not using these specific Microsoft AI cloud services are unlikely to be impacted.

4. What to do now

  • Check with your IT provider if your organisation uses Azure AI Foundry or related Microsoft AI cloud services.
  • Ensure that all Microsoft Azure services are updated with the latest security patches from Microsoft.
  • Review access controls and monitor for any unusual activity on your cloud accounts.
  • Ask your IT provider to confirm that authentication settings are correctly configured to prevent unauthorized access.

5. Ask your IT provider

Can you confirm if we use Microsoft Azure AI Foundry services, and have all critical security patches been applied to protect against privilege escalation vulnerabilities?

6. Bottom line

If you use Microsoft Azure AI services, make sure they are fully updated to avoid serious security risks.

Information based on NVD, CISA KEV, and reputable security news reporting.

Back to Vulnerability Briefs