Skip to the content
AI and Cybersecurity Learning and Practice Center Test automation Open the Learning Center

IncidentAI-enabled attacks

Storm-0940: ARTEX AI agents and hijacked ConnectWise app identities reached for more than 2,250 Microsoft 365 tenants

CriticalConnectWiseMicrosoft

Microsoft Security Research and Microsoft Threat Intelligence published on October 21, 2026, and updated on October 23, their assessment that Storm-0940, a China-state aligned actor active since 2021 that usually sprays passwords, exploits internet-facing applications and adds credentials to existing OAuth applications, was responsible for the recently disclosed ConnectWise compromise. It began using ARTEX, an open-source multi-agent penetration testing framework that can use any AI model, as early as September 2026, for scanning and reconnaissance, password spraying and the abuse of OAuth applications at downstream victims. The group gained unauthorized access related to ConnectWise Cloud Backup and took control of the identities of several ConnectWise cloud applications, which its agents used to authenticate to customers' Microsoft 365 tenants without interactive sign-in, at more than 2,250 tenants. Most targeting failed, but it reached privileged cloud persistence at many organizations, in multiple cases less than three hours after targeting began, and followed with application credential changes through Microsoft Graph, privilege changes including a self-assigned Global Administrator role in one case, and bulk reads of Azure Key Vault secrets. Its spray from October 10 to 14 made more than 151,000 failed attempts against 327 accounts at eight organizations, peaking at 2,900 a minute on October 11, with no successful sign-in. Public reports surfaced on October 15 and ConnectWise published its first advisory on October 16. ConnectWise shut down its Cloud Backup, SaaS Security and Cloud Management services; its chief executive said containment does not remove access already gained in customer tenants. Microsoft has not said how many organizations were compromised and notes that a sign-in does not by itself mean data was taken.

Replay the attack step by step in the Learning Center

How it happened

  1. Gain access related to ConnectWise Cloud Backup and take control of the identities of ConnectWise cloud applications that customers had allowed into their Microsoft 365 tenants.
  2. Let ARTEX's AI agents use those trusted identities to sign in to tenants at machine speed, without any user login, alongside password spraying.
  3. Where a sign-in works, add application credentials through Microsoft Graph, raise privileges and read Key Vault secrets, so access outlives the vendor's cleanup.

Impact

Privileged, persistent cloud access at many of the organizations targeted, in some cases within three hours; Microsoft has not said how many were compromised or whether data was taken.

The fix

Keep the affected ConnectWise application connections disabled until ConnectWise gives formal restoration instructions, and preserve configurations and logs before deleting anything. Find the applications by application ID, not display name, review their successful and failed service-principal sign-ins from at least October 15 (October 1 where records exist), and audit credential additions, role changes, consent grants and mailbox and file activity; then revoke unneeded permissions and rotate credentials. Give multi-tenant apps the least privilege, and alert on new app credentials and privileged role assignments.

Lessons

  • A third-party app with standing permissions in a tenant can be used without anyone's password; its access is only as safe as the vendor.
  • Agentic attack tools cut the time from target to privileged access to hours, so detection has to watch app credential and role changes, not only user sign-ins.
  • Switching off the vendor's connection does not evict an attacker who has already planted access elsewhere in the tenant.

Sources