China-Linked Hackers Leverage AI Agents in Widespread ConnectWise Breach

For many IT administrators and business owners, the recent weeks have been defined by a heightened sense of urgency regarding cloud security. A large-scale supply chain attack targeting ConnectWise has forced organizations to scrutinize their Microsoft 365 environments, as a China-state-aligned group utilized artificial intelligence to accelerate its intrusion. The incident highlights a shift toward faster, more automated threats that leave traditional defensive teams with minimal time to react. For ordinary users in affected companies, this may mean temporary service disruptions or increased password resets as firms verify the integrity of their data.

The primary driver of this attack was the deployment of ARTEX, an open-source framework that employs large language model-driven agents to conduct autonomous penetration tests. Microsoft’s security team reported that these AI agents significantly sped up reconnaissance and credential abuse. The system operated continuously with minimal human intervention, allowing the attackers to scan systems and attempt logins at a pace that overwhelmed standard detection methods. This technological edge reduced the time between initial targeting and privileged access to under three hours in several instances.

The scale of the operation was substantial, with the group attempting sign-ins to more than 2,250 Microsoft 365 tenants using stolen ConnectWise application credentials. These credentials allowed the attackers to authenticate as apps rather than individual users, bypassing typical interactive login checks. Additionally, automated password spraying targeted 327 accounts across eight organizations, resulting in over 151,000 failed login attempts. While most of these brute-force attempts were locked out, the use of valid app credentials proved far more effective in gaining a foothold in customer environments.

At the peak of the activity on October 11, the password spraying campaign reached 2,900 attempts per minute, totaling more than 111,000 failures in a single day. In contrast, the app-based sign-ins were more precise and devastating. Attackers used just four application credentials to generate over 200 successful sign-ins in under four minutes. Once inside, the agents performed directory operations, changed app credentials, and modified privileged roles. Security firm Huntress observed one instance where an app assigned itself the top administrator role, a behavior that triggered specific alerts for self-assigned global administrator permissions.

The group, designated as Storm-0940, has been active since 2021 and targets industries including aerospace, defense, semiconductors, government, and think tanks. Microsoft clarified that the breach did not stem from a vulnerability in its own systems but rather from unauthorized access to the multi-tenant ConnectWise application. The activity was characterized by a systematic sweep of tenants in sorted order, suggesting the use of a script working from a predefined target list. By October 16, Huntress had confirmed unauthorized sign-ins for 162 tenants, a number that rose to nearly 1,000 by the following day.

This incident is part of a broader trend of threat actors in various regions adopting AI-driven tools for cyber operations. Public reports this month have linked similar breaches at South Korean financial institutions to the use of ARTEX, with suspects allegedly based in China. In response to the growing adoption of the framework, its developer, known by the handle Autumn-27, announced that the project would cease updates and transition to closed source. This move may impact the availability of the tool for other potential attackers in the future.

ConnectWise has suspended several services, including Cloud Backup and SaaS Security, as a precautionary measure while it conducts a phased restoration. The company expects recovery to take several days and has not found indicators that partner backup data was compromised. Microsoft advises customers to keep affected app connections disabled until formal restoration instructions are issued and to review recent activity logs. The use of AI in this supply chain attack marks a significant evolution in threat tactics, emphasizing the need for organizations to adapt their defensive strategies to keep pace with automated, high-speed intrusions.