Threat actors use ARTEX AI to steal Azure Key Vault secrets
Original title: Hackers Use ARTEX AI to Steal Azure Key Vault Secrets Just 5 Minutes After Scanning GitHub
Microsoft says one agent went from GitHub searches to Key Vault access in five minutes, using stolen service principals.
Microsoft Threat Intelligence says criminals and state-linked groups are using the ARTEX penetration-testing framework to find secrets in public GitHub code and sign in as Azure apps. From August 2026 the actors searched for company names, secret-related terms and cloud or AI keys. From October 28, Microsoft saw successful service-principal sign-ins, including from a compromised vendor multitenant app that reached downstream customers. Agents listed cloud resources, pulled credential-bearing settings, downloaded SharePoint and OneDrive files, retried failed API calls and used Azure Run Command on virtual machines.
Unconfirmed: Todyl’s counts are described as minimums, and Microsoft says 199.19.77.18 may include unrelated Tor traffic.
- Todyl reports about 45 app registrations and access to 27 organizations.
- Todyl cites 81 downloaded files and about 2,279 mailbox items as a minimum.
- Sign-in bursts reached 24 in 10 seconds; one Key Vault retry succeeded in about 28 seconds.
- Microsoft named 199.19.77.84, 65.87.223.26 and 157.22.126.128, plus Tor exit 199.19.77.18.
- One partner saw thousands of dollars in AI usage charges in about 10 minutes.
Impact: Stolen app secrets used to read vaults, files and cloud credentials.
- 1.Treat any secret that was ever public as stolen, rotate it and review its use.
- 2.Block 199.19.77.84/32 only, and review Entra sign-ins back to at least October 28.
- 3.Move apps to managed identities and enable GitHub secret scanning and push protection.
- ATT&CK
- T1552.001 Credentials In FilesT1078.004 Cloud AccountsT1528 Steal Application Access TokenT1530 Data from Cloud Storage
- Vendors
- Microsoft
- Malware
- ARTEX
- Organizations
- Microsoft Threat IntelligenceTodyl
Indicators of compromise
auto-extracted · verify before use · export allAll| Type | Indicator | Context |
|---|---|---|
| ipv4 | 157.22.126.128 | r agent and attempts to get App Service tokens. The address 157.22.126[.]128 scouted AI workloads on Azure AI services. Microsoft also |
| ipv4 | 199.19.77.18 | ed AI workloads on Azure AI services. Microsoft also listed 199.19.77[.]18 as a Tor exit node that may bring unrelated noise. Our co |
| ipv4 | 199.19.77.84 | w successful service principal sign-ins from the IP address 199.19.77[.]84 across several application IDs. A service principal is th |
| ipv4 | 65.87.223.26 | seconds. Microsoft named two more IP addresses. The address 65.87.223[.]26 showed shell commands hidden in a user agent and attempts |
From the original article
4 min read at cybersecuritynews.com
Microsoft says threat actors are using an AI tool called ARTEX to find secrets in public GitHub code and use them against cloud accounts. In one case, the AI agent went from searching GitHub to pulling data from Azure Key Vault in only five minutes. ARTEX AI Azure Key Vault secrets Searching public code for […]
By Kavichselvan · cybersecuritynews.com
A short excerpt; the article and its images belong to the publisher. Summary, key facts and assessment above are AI-written from the article and can be wrong.