Hackers Use AI Agents to Break Into Thousands of Microsoft Accounts

A sophisticated cyberattack linked to China is raising questions about the speed of automated threats, after hackers deployed artificial intelligence agents to breach thousands of Microsoft 365 tenants. Microsoft reported in a Security blog that a state-aligned actor known as Storm-0940 utilized an open-source tool called ARTEX to execute a large-scale supply chain attack against ConnectWise. Following unauthorized access to a multi-tenant application, the group attempted sign-ins at more than 2,250 tenants using stolen credentials. In some instances, the attackers were able to reach privileged access in less than three hours from the moment of initial targeting.

The breach originated from ConnectWise, a software provider, which noted the unauthorized access and suspended specific services including Cloud Backup, formerly SkyKick, as a precaution. This action caused a disruption to partner backup data, though ConnectWise stated there are no indicators to date that the data was compromised. A phased restoration process has begun, with recovery expected to take several days. Microsoft noted that the activity did not result from a compromise of or a vulnerability in Microsoft’s systems, products, or services, but rather from the stolen credentials associated with the ConnectWise applications.

The use of ARTEX, described as a system of LLM-driven agents that run an autonomous penetration test, facilitated rapid scanning, reconnaissance, and password spraying. Microsoft stated these agents can work around the clock with little human interaction, operating at a pace that leaves defenders with little time to detect and respond. While automated password guessing targeted 327 accounts across eight organizations with over 151,000 failed logins, the group successfully authenticated as applications without interactive user sign-in. It remains uncertain how the ConnectWise credentials were initially stolen or how many tenants were actually compromised.

Storm-0940, active since 2021, is identified as a China-state aligned actor with no known industry aliases. The group targeted industries including aerospace and defense, semiconductors, government, and think tanks, with 14 sectors listed in total. Security firm Huntress observed suspicious activity on October 15, where an application added itself to the top administrator role within a tenant. By October 16, Huntress confirmed unauthorized sign-ins for 162 tenants, and the following day, almost 1,000 Entra tenants showed sign-in activity from the attacking IP address.

The activity involved planting new secrets set to expire in 2028 and making changes to identities, groups, and privileged roles. Microsoft advises customers to keep affected app connections disabled until formal restoration instructions are issued and to review activity from this month where logs allow. The use of such tools is not isolated to this incident; public reporting this month tied ARTEX to breaches at South Korean financial institutions, where the suspect was likely based in China and allegedly used the framework with Anthropic’s Claude Code. The developer of ARTEXT, known by the handle Autumn-27, stated on GitHub that the project will no longer be updated and will become closed source.

The attack’s sweep through tenants in sorted order, occurring only seconds apart, indicated a script working from a target list. While the use of AI agents represents a departure from traditional attacks, the underlying compromise stems from the supply chain breach at ConnectWise. As restoration continues, the immediate priority remains disabling affected app connections until official instructions are provided.