Microsoft has said a China-state-aligned threat actor used an artificial intelligence (AI)-driven penetration-testing framework to automate cyberattacks targeting hundreds of organisations worldwide, with some attacks resulting in privileged access to cloud infrastructure in less than three hours.
In a security report, Microsoft Threat Intelligence identified the actor as Storm-0940 and said it began using the open-source ARTEX framework as early as September 2026 to automate reconnaissance, password spraying and the abuse of compromised applications. ARTEX can use AI models to automate a range of offensive security workflows.
Microsoft assessed that Storm-0940 was responsible for the recently disclosed ConnectWise supply-chain compromise involving ConnectWise Cloud Backup, a multi-tenant application. The report said attackers acquired control of identities associated with multiple ConnectWise cloud applications and used stolen application credentials to authenticate to customer environments.
The company said AI agents enabled attackers to operate around the clock and pursue multiple attack paths simultaneously. During the ConnectWise incident, Storm-0940 used ARTEX to accelerate application abuse and attempt to access customers’ Microsoft 365 environments through applications that had already been granted access, without requiring an interactive user sign-in.
The attacks targeted organisations across sectors including aerospace and defence, automotive, energy, financial services, government, manufacturing, semiconductors, software and technology, as well as think tanks and non-profits. Although most targeting attempts were unsuccessful, Microsoft said Storm-0940 achieved privileged cloud persistence at many organisations.
Following the ConnectWise compromise, the attackers attempted to sign in to more than 2,250 customer tenants, generating approximately 2,900 sign-in attempts per minute, Microsoft said.
Between October 10 and 14, 2026, Microsoft recorded more than 151,000 failed logon events in a password-spraying campaign targeting 327 accounts across eight organisations. No successful authentication, token issuance, Microsoft Graph call or directory change was observed during that campaign.
Microsoft cautioned that the presence of an affected application or successful authentication does not necessarily mean customer data was accessed or exfiltrated.
The company advised organisations to keep affected ConnectWise application connections disabled until the vendor provides formal restoration instructions. It also recommended preserving investigation evidence and reviewing application sign-ins, credential changes, permission grants, privileged role assignments and Microsoft Graph activity.
Also Read: Microsoft, Council of Europe Sign AI Framework on Human Rights and Democracy
