We’ve talked for a while about attackers using AI. Microsoft’s latest threat report shows what that looks like in practice, and it’s fast.
Microsoft Threat Intelligence says Storm-0940, a China-state aligned group active since 2021, has been using ARTEX, an open-source agentic AI penetration testing framework, to automate attacks against hundreds of organisations worldwide since as early as September 2026. Microsoft also assesses that Storm-0940 was behind the recently disclosed supply-chain compromise of ConnectWise Cloud Backup.
ARTEX is a system of AI agents that can run a penetration test from reconnaissance through to data exfiltration with little human input. It’s built to work with any AI model. Microsoft says it has seen threat actors in multiple regions using it.
The scale stands out. After the ConnectWise compromise, Storm-0940 used ARTEX to try to sign in to more than 2,250 Microsoft 365 tenants. In an earlier password-spray campaign between 10 and 14 October, Microsoft recorded more than 151,000 failed logon events, peaking at 2,900 failures per minute on 11 October. That spray didn’t succeed, but other attacks did: in several cases, the time from first targeting to privileged cloud access was under three hours.
The trick was trust. By taking over the identities of ConnectWise cloud applications that customers had already granted access to, the attackers’ AI agents could authenticate into customer environments without anyone signing in. In some tenants they went on to run reconnaissance through Microsoft Graph, add application credentials for persistence and try to change privileged roles.
Microsoft stresses that this didn’t come from a flaw in its own systems, and that an affected app or a successful sign-in doesn’t automatically mean data was stolen.
Its advice for affected customers includes keeping the ConnectWise app connections disabled until ConnectWise provides restoration instructions, preserving logs before deleting anything, reviewing activity from at least 15 October, and checking for unexpected credential, role and consent changes.
Why it matters
AI agents don’t sleep, and they can try many attack paths at once. That shrinks the time defenders have to react from days to hours. The lesson for every business using Microsoft 365 or Google Workspace: review which third-party apps have access to your tenant, remove the ones you don’t need, and keep permissions as narrow as possible. A trusted integration with too much access is now one of the fastest ways in.
Sources: Microsoft Security Blog. Written by the xRapid Media LLP team.
Selling to the IT market? xRapid Media LLP is a High Tech B2B Marketing company. We help technology companies reach decision-makers with verified B2B data, account-based campaigns, email and LinkedIn outreach and appointment setting. Call +91 9767196309 or book a free B2B marketing consultation.