Microsoft Warns of AI-Powered Attacks Linked to ConnectWise Compromise
Microsoft says attackers used AI to move from reconnaissance to privileged cloud access in a matter of hours, giving defenders far less time to respond.
Key Takeaways:
- Microsoft observed Storm-0940 using the ARTEX AI framework to automate cloud attack activities.
- The campaign leveraged a compromised ConnectWise application to target thousands of downstream tenants.
- Microsoft recommends tighter governance of third-party applications, stronger monitoring, and least-privilege access controls.
A newly observed campaign by Microsoft Threat Intelligence shows how the China-aligned threat actor Storm-0940 used the open-source AI penetration testing framework ARTEX to automate and accelerate attacks, including activity linked to the recent ConnectWise supply chain compromise. This campaign shows how agentic AI can compress attack timelines from days or weeks into just a few hours.
Organizations increasingly trust third-party cloud applications with broad permissions to access Microsoft 365 and other business services. After compromising a multi-tenant ConnectWise application, Storm-0940 leveraged the AI-powered ARTEX platform to automate reconnaissance, password-spraying, credential abuse, and cloud privilege escalation across thousands of downstream tenants. This framework enabled attacks at machine speed, which allowed the threat actor to attempt multiple intrusion paths simultaneously and quickly identify opportunities for compromise.
Attackers moved through cloud environments at machine speed
The combination of a trusted third-party application and autonomous AI agents significantly increased the scale and speed of the campaign. Microsoft observed the attackers attempting sign-ins against more than 2,250 tenants and, in several cases, gaining privileged cloud access in under three hours.
Once authenticated through trusted applications, the attackers conducted reconnaissance through Microsoft Graph, modified application credentials to maintain persistence, and attempted privilege changes that could expand access across victim environments. Even when detection tools generated alerts, security teams often had little time to investigate and respond before the attackers progressed further into the environment.
Microsoft recommends stronger OAuth and application governance
Microsoft recommends treating third-party applications as high-value security assets and conducting immediate reviews of affected application permissions, service-principal sign-ins, audit logs, and credential changes. Organizations should apply least-privilege access to OAuth applications, monitor for unexpected privilege assignments or credential additions, investigate anomalous Microsoft Graph activity, and verify that vendor remediation efforts fully remove unauthorized access. Microsoft also advises using tools such as Microsoft Defender XDR, Microsoft Defender for Cloud Apps, Microsoft Entra, and the new Security Dashboard for AI to detect and investigate AI-enabled threats more effectively.
This incident highlights that AI-powered tools can improve legitimate security testing and streamline defensive operations, but the same capabilities can dramatically enhance attacker efficiency. As AI lowers the cost and effort required to conduct reconnaissance, credential attacks, and privilege escalation, organizations gain productivity benefits from AI adoption while also inheriting a much faster threat landscape. Security teams invest in stronger monitoring, tighter application governance, and faster response capabilities, accepting additional operational complexity in exchange for resilience against increasingly automated attacks.
Rabia has a master's degree in Software Engineering and she has years of experience writing professionally about Microsoft products and other technologies. Rabia has also written for OnMSFT.com as well as Windows Report. She is always up to date on t...
