Storm-0940 and ARTEX: How China's AI-Powered Agents Hijacked ConnectWise and 2,250 Microsoft 365 Tenants
Executive Summary
In October 2026, Microsoft attributed a sophisticated supply-chain attack against ConnectWise's Cloud Backup service to Storm-0940, a China-state aligned threat actor active since 2021. The group leveraged ARTEX, an open-source agentic AI penetration testing framework, to automate password-spraying at 2,900 sign-in attempts per minute, ultimately targeting more than 2,250 Microsoft 365 tenants across eight organizations. AI agents autonomously progressed from initial authentication through reconnaissance, OAuth abuse, credential modification, and privilege escalation with minimal human interaction. ConnectWise temporarily shut down services following the attack, and Microsoft confirmed evidence of follow-on persistence activity in multiple customer environments.
This incident marks a critical inflection point in nation-state tradecraft: the weaponization of agentic AI frameworks enables adversaries to compress attack timelines from days to minutes, outpacing human defenders. The combination of supply-chain entry, AI-driven automation, and OAuth abuse signals a new era of high-velocity, low-visibility intrusions that demand equally automated defensive countermeasures.
Why This Matters Now
Agentic AI tools like ARTEX are now openly available and adopted by multiple nation-state and criminal groups, enabling any moderately resourced adversary to conduct enterprise-scale supply-chain intrusions autonomously. With ConnectWise serving thousands of MSPs and enterprises globally, the blast radius of a single compromised backup integration is enormous—and defenders have minutes, not hours, to respond.
Attack Path Analysis
Storm-0940, a China-state aligned threat actor, leveraged the ARTEX agentic AI penetration testing framework to conduct a high-velocity password-spraying campaign against Microsoft 365 and Exchange Online accounts, then pivoted through the compromised ConnectWise Cloud Backup supply-chain integration to access downstream customer tenants. Once inside, AI agents autonomously performed reconnaissance, modified OAuth application credentials, and manipulated privileged identity roles to establish persistent backdoors, with follow-on activity indicating potential for data exfiltration and sustained access across hundreds of organizations globally.
Kill Chain Progression
This analysis maps confirmed threat intelligence to the full cloud kill chain to show where defensive gaps would emerge as an attack progresses.
Initial Compromise
Description
Storm-0940 deployed ARTEX-driven AI agents to execute a massive password-spraying campaign against 327 Microsoft 365 and Exchange Online accounts across eight organizations, generating over 151,000 failed logon attempts peaking at 2,900 per minute. The threat actor then exploited the privileged trust relationship of ConnectWise Cloud Backup's integration with customer M365 environments, using the compromised third-party application as a supply-chain entry vector into over 2,250 downstream tenants.
MITRE ATT&CK® Techniques
Compromise Software Supply Chain
Password Spraying
Application Access Token
Additional Cloud Roles
Cloud Account Discovery
Steal Application Access Token
Command and Scripting Interpreter
Exfiltration Over Web Service
Potential Compliance Exposure
Mapping incident impact across multiple compliance frameworks.
NIST SP 800-53 Rev. 5 – Authenticator Management and Account Control
Control ID: IA-5, IA-8, AC-2
CISA Zero Trust Maturity Model 2.0 – Continuous Identity Verification and Least Privilege Enforcement
Control ID: Identity Pillar – Authentication & Authorization
DORA (Digital Operational Resilience Act) – Third-Party ICT Risk Management and Incident Reporting
Control ID: Article 19 – ICT-related Incident Reporting; Article 28 – ICT Third-Party Risk
NIS2 Directive – Supply Chain Security and Significant Incident Reporting
Control ID: Article 21 – Cybersecurity Risk Management Measures; Article 23 – Reporting Obligations
PCI DSS 4.0 – Multi-Factor Authentication and Third-Party Service Provider Controls
Control ID: Requirement 8.4 – MFA; Requirement 12.8 – Third-Party Risk
NYDFS 23 NYCRR 500 – MFA Enforcement and Third-Party Cybersecurity Controls
Control ID: Section 500.12 – Multi-Factor Authentication; Section 500.11 – Third-Party Service Provider Security
Sector Implications
Industry-specific impact of the vulnerabilities, including operational, regulatory, and cloud security risks.
Information Technology/IT
ConnectWise supply-chain compromise exposed IT managed service environments to Storm-0940's ARTEX AI agents, enabling OAuth abuse, credential modification, and lateral movement across Microsoft 365 tenants at scale.
Financial Services
Explicitly targeted by Storm-0940, financial sector organizations face PCI.4.0 compliance gaps as AI-automated password-spraying and privilege escalation threaten Microsoft 365 environments requiring Zero Trust segmentation and egress controls.
Defense/Space
Aerospace, defense, and military organizations named as direct Storm-0940 targets face NIST.800-53 compliance exposure as ARTEX-driven reconnaissance and persistent backdoor establishment threaten classified operational environments.
Think Tanks
Think tanks explicitly listed among Storm-0940's targeted sectors face heightened exfiltration risk, with unencrypted M365 traffic and insufficient east-west segmentation enabling China-state aligned actors to harvest sensitive policy intelligence.
Sources
- Microsoft pins ConnectWise supply-chain attack to China-state aligned threat grouphttps://cyberscoop.com/microsoft-attributes-connectwise-supply-chain-attack-china-storm-0940/Verified
- Microsoft Security Blog: Storm-0940 and ARTEX AI-driven attackshttps://www.microsoft.com/en-us/security/blog/Verified
- ConnectWise Security Advisory: Cloud Backup Service Incidenthttps://www.connectwise.com/company/trust/security-bulletinsVerified
- CISA Known Exploited Vulnerabilities Cataloghttps://www.cisa.gov/known-exploited-vulnerabilities-catalogVerified
- Microsoft Threat Intelligence: Nation-State Actor Profileshttps://learn.microsoft.com/en-us/defender/threat-intelligence/threat-actor-namingVerified
Frequently Asked Questions
Cloud Native Security Fabric Mitigations and ControlsCNSF
Based on the attack progression modeled above, these are the defensive controls that would constrain each stage.
Aviatrix Zero Trust CNSF is directly relevant to this incident because Storm-0940 exploited trusted supply-chain integration pathways and high-privilege cloud API access to move freely across hundreds of M365 tenants. By enforcing workload-level segmentation and restricting lateral reachability between integrated cloud services, CNSF would likely have reduced the attacker's blast radius well before tenant-wide persistence was established.
Control: Cloud Native Security Fabric (CNSF)
Mitigation: CNSF fabric-level controls would likely constrain the reachability of compromised supply-chain integration paths across downstream tenant environments, reducing how far an initial foothold through the ConnectWise service account could propagate into protected workload boundaries.
Control: Zero Trust Segmentation
Mitigation: Zero Trust segmentation policies would likely constrain the scope of privilege escalation by limiting which cloud workloads and identity planes a compromised service account context could interact with, reducing the attacker's ability to freely traverse from the integration layer into high-privilege tenant administration.
Control: East-West Traffic Security
Mitigation: East-west traffic security controls would likely reduce the attacker's ability to freely enumerate and pivot across connected cloud services and directory objects, constraining how much of the tenant structure could be explored from any single compromised workload or identity context.
Control: Multicloud Visibility & Control
Mitigation: Multicloud visibility and control capabilities would likely improve detection of anomalous outbound API communication patterns originating from cloud workloads, reducing the attacker's ability to sustain undetected ARTEX agent orchestration hidden within legitimate-looking SaaS traffic flows.
Control: Egress Security & Policy Enforcement
Mitigation: Egress security and policy enforcement would likely constrain the outbound data paths available to ARTEX agents, reducing the volume and variety of sensitive M365 directory data and credentials that could be staged and moved to attacker-controlled infrastructure from compromised tenant environments.
Had earlier CNSF controls meaningfully constrained supply-chain propagation, lateral movement, and egress paths, the residual impact across the 2,250 targeted tenants would likely be reduced in scope, with fewer organizations exposed to persistent OAuth backdoors and fewer privileged roles available for sustained attacker access.
Impact at a Glance
Affected Business Functions
- Microsoft 365 Identity and Access Management
- Cloud Backup and Data Protection Services
- Privileged Role and Directory Administration
- Third-Party Managed Service Provider (MSP) Operations
Estimated downtime: 7 days
Estimated loss: N/A
Threat actor AI agents authenticated to Microsoft 365 and Exchange Online environments across more than 2,250 tenants spanning industries including aerospace and defense, financial services, government, energy, semiconductors, and technology. Exposed data categories include Microsoft 365 directory objects, identity and privileged role assignments, OAuth application credentials, Exchange Online mailbox contents, and potentially sensitive corporate communications and documents. In environments where follow-on activity was confirmed, persistent backdoor access was established via credential modification, posing ongoing risk of exfiltration of intellectual property, regulated financial data, government communications, and PII at enterprise scale. Financial loss is not yet publicly confirmed but is assessed as potentially significant given the breadth of affected sectors and tenant count.
Recommended Actions
Key Takeaways & Next Steps
- • Deploy Zero Trust Segmentation to enforce least-privilege OAuth and API boundaries: CNSF Zero Trust Segmentation with identity-based policy controls would restrict supply-chain application integrations like ConnectWise Cloud Backup to narrowly scoped, explicitly authorized API permissions, preventing AI agents from abusing over-privileged OAuth applications to escalate into tenant directory and role management operations.
- • Activate Multicloud Visibility & Control to detect anomalous authentication velocity and agentic API behavior: CNSF centralized traffic observability and anomaly detection capabilities should be configured to baseline normal M365 Graph API call patterns and alert on the high-velocity, automated interaction signatures characteristic of ARTEX-driven agents — such as 2,900 sign-in attempts per minute and rapid sequential directory enumeration — providing defenders early warning before persistence is established.
- • Enforce Egress Security & Policy Enforcement to block unauthorized outbound data flows from M365 environments: CNSF egress filtering with FQDN-level policy enforcement and outbound traffic controls should be applied to cloud workload egress paths to prevent ARTEX-orchestrated exfiltration of directory data, email content, and OAuth tokens to attacker-controlled infrastructure, aligned with ZTMM.Data and NIST.800-53.SC-18 requirements.
- • Leverage East-West Traffic Security controls to limit lateral pivot scope across tenant services: Implementing CNSF east-west traffic security between M365 service integrations, workload-to-workload API flows, and inter-tenant communication channels would contain the blast radius of a compromised third-party application, preventing Storm-0940 agents from freely pivoting across directory services, Exchange Online, and downstream connected applications once initial access was obtained.
- • Instrument Cloud Firewall (ACF) with AI-driven traffic discovery to identify and block malicious infrastructure reuse: CNSF cloud firewall capabilities with AI-assisted egress traffic discovery should be used to identify and block the malicious infrastructure Storm-0940 reused across both the password-spray and ConnectWise supply-chain phases, enabling proactive denial of command-and-control and exfiltration channels before downstream tenant compromise progresses to persistent access or data loss.
