Microsoft put a name on the ConnectWise hack Wednesday, and it’s a grim milestone for defenders. The company said Storm-0940, a China state-aligned hacking crew, was behind the recent supply-chain compromise of the ConnectWise Cloud Backup application — and ran its follow-on attacks through ARTEX, an agentic AI penetration-testing framework that did the break-in work at machine speed.

The numbers are the story. Since September, Storm-0940 targeted hundreds of organizations worldwide, with attack attempts touching more than 2,250 Microsoft tenants at peaks of 2,900 sign-in attempts per minute. In several cases, the time from first targeting to privileged access on a victim’s cloud infrastructure was under three hours. Nobody had to type a password.

How the Microsoft ConnectWise attack worked

Storm-0940 started with what Microsoft calls the supply-chain compromise of a multi-tenant ConnectWise Cloud Backup application. From there, the group seized the identities of multiple ConnectWise cloud applications — including, by at least one account, a compromised API secret — and used them to walk straight into customer environments. ConnectWise temporarily pulled the plug on its Cloud Backup, SaaS Security and Cloud Management services after discovering the unauthorized access to application identities, The Register reported.

Here’s the part that should worry every Microsoft 365 administrator: the attackers authenticated through applications that already had access, so no interactive user sign-in ever happened. Because OAuth application credentials authenticate as the app rather than a person, multi-factor authentication never fired. The trusted app was the key, and it already fit the lock.

Once inside, the AI agents got to work through Microsoft Graph. Microsoft observed reconnaissance, application credential changes, and attempts to modify privileges that could enable expanded or persistent access — including pushes toward global administrator rights. The company was explicit that none of this came from a flaw in its own systems: “the activity conducted by the threat actor through the usage of ARTEX framework did not result from a compromise of or a vulnerability in Microsoft corporate systems, products, or services.”

AI at machine speed

ARTEX is the other headline here. It’s a Go-based agentic AI penetration-testing framework that runs the full attack chain — reconnaissance, vulnerability scanning, password spraying, OAuth abuse, credential changes, privilege escalation, even data-exfiltration attempts — almost autonomously, with minimal human involvement. It works with multiple language models, and Microsoft says the attackers let the agents attempt multiple breach paths simultaneously.

“Agents are not constrained by factors affecting humans — they can work around the clock, on weekends, and on holidays and attempt multiple breach paths simultaneously,” Microsoft wrote in its report. That single line is doing a lot of work. It also describes exactly what defenders saw.

Between October 10 and 14, infrastructure at a single IP address ran an automated cross-tenant password-spray campaign targeting 327 accounts across eight organizations, generating more than 151,000 failed logon events. Activity peaked on October 11 at 2,900 failures per minute — 111,000 failed logons in a single day across more than 2,250 tenants, according to coverage of Microsoft’s findings. That spray campaign didn’t produce successful logins, but it wasn’t the main event: on October 15, the agents had obtained the identities of multiple ConnectWise cloud applications and used them to authenticate into customer environments instead.

The victims

Microsoft says the targets span aerospace and defense, automotive, chemicals, energy, financial services, government, legal, manufacturing, media, military, semiconductor, software, technology, and think tanks and nonprofits. Huntress Security, a third-party security firm, independently confirmed roughly 300 compromised tenants with matching timelines — a second data point corroborating Microsoft’s attribution.

Storm-0940 is no newcomer. Microsoft first detailed the group in November 2024, when it was running highly evasive password-spray attacks through a botnet called Quad7 (also tracked as CovertNetwork-1658). The company assesses the actor has been active since at least 2021. ARTEX itself has a pedigree in misuse: the framework was made closed-source after being abused in cyber attacks against South Korean financial institutions, yet it clearly kept circulating.

Why this matters

Two uncomfortable shifts are on display. First, the OAuth abuse angle: the victims’ weakness wasn’t a missing patch, it was the architecture of trusted third-party apps with broad cross-tenant permissions. When an attacker holds a legitimate application’s credentials, the sign-in looks legitimate — no phishing email to catch, no MFA prompt to deny.

Second, the clock. A three-hour window from targeting to privileged access collapses the response time defenders plan around. Incident-response playbooks assume humans on the other side, taking weekends off and triaging alerts in business hours. Storm-0940’s agents don’t sleep, and they don’t do one target at a time.

Microsoft published indicators of compromise, Defender detections, and guidance for investigating service-principal sign-ins and unexpected application credential or privilege changes alongside the disclosure. If you run Microsoft 365 with ConnectWise integrations, this one is worth reading directly.

Frequently asked questions

What is the Microsoft ConnectWise hack?

Microsoft says China state-aligned Storm-0940 used the ARTEX agentic AI penetration-testing framework to exploit a ConnectWise Cloud Backup supply-chain compromise, then attacked hundreds of organizations’ Microsoft 365 environments at machine speed.

Who is Storm-0940?

Storm-0940 is a China state-aligned threat actor Microsoft first detailed in November 2024. It has been active since at least 2021 and previously used the Quad7 botnet for evasive password-spray attacks.

What is ARTEX?

ARTEX is a Go-based agentic AI penetration-testing framework that can run reconnaissance, scanning, password spraying and exploitation workflows almost autonomously. It was made closed-source after being abused in attacks on South Korean financial institutions.

How many organizations did the Storm-0940 AI attack hit?

Microsoft logged attack attempts against more than 2,250 tenants, with peaks of 2,900 sign-in attempts per minute. A separate October 10–14 password-spray campaign produced 151,000 failed logons, and Huntress Security independently confirmed about 300 compromised tenants.

Sources: Microsoft Threat Intelligence, CyberSecurityNews, AI Weekly, The Register