A China state-aligned hacking group used AI agents to break into roughly 2,250 organizations through a breach of IT management vendor ConnectWise, according to Microsoft, which says the attack’s machine speed is giving defenders “little time to detect and respond.”
Microsoft’s security research and threat-intelligence teams published their findings in a report on Wednesday, naming the group Storm-0940, which the company assesses as aligned with the Chinese state. The hackers broke into ConnectWise’s cloud backup service and stole application credentials — the credentials ConnectWise uses to establish data links with its corporate customers. With those credentials in hand, Microsoft said, the attackers could log in to a customer’s Microsoft-hosted environment as if they were an authorized user, gaining access to emails and other files.
What set this operation apart, according to Microsoft, was the tooling: ARTEX, a Go-based agentic AI penetration-testing framework that pairs AI models with automated hacking techniques normally used to stress-test cyber defenses. Storm-0940 used it to speed up scanning and reconnaissance, password spraying, and OAuth application abuse against downstream victims, Microsoft said.
“Storm-0940’s use of ARTEX accelerated the speed of compromised application abuse to authenticate to target environments through applications that had already been granted access, allowing the threat actor to operate without requiring an interactive user sign-in,” Microsoft said in the report.
The numbers behind the campaign are stark. The hackers targeted more than 2,250 Microsoft 365 tenants at a rate of almost 3,000 attempted sign-ins per minute, Microsoft said. In several cases, the attackers reached privileged access to an organization’s cloud infrastructure less than three hours after initial targeting. AI-accelerated follow-on activity included reconnaissance, application credential changes, and attempts to modify privileges that could extend or preserve access.
ARTEX is not new to real-world abuse. The same framework was used in a recent compromise of South Korea’s financial systems, many of which had weak security and lacked multi-factor authentication, according to TechCrunch. Its developers made ARTEX closed-source after that abuse — a reminder that the AI pentesting tools maturing inside the legitimate security industry are maturing in the hands of attackers at the same time.
The breach did not start with the ConnectWise credentials. Between October 10 and 14, Microsoft said, infrastructure at IP address 144.172.64[.]99 ran an automated cross-tenant password-spray campaign that generated more than 151,000 failed sign-in events against 327 accounts at eight organizations. Activity peaked on October 11 at 2,900 failed attempts per minute, and 48 accounts each received more than 1,000 failed attempts. None of those attempts succeeded, Microsoft said — and the attackers then pivoted to the stolen ConnectWise application identities.
Storm-0940 is no newcomer. Microsoft assesses the group has been active since at least 2021 and first detailed it in November 2024, when it was relying on a botnet known as Quad7, or CovertNetwork-1658, for highly evasive password-spray attacks. Since September 2026, Microsoft said, the group has targeted hundreds of organizations across more than 15 sectors, including aerospace and defense, automotive, chemicals, energy, financial services, government, legal, manufacturing, media, military, semiconductors, software, and technology.
ConnectWise said it has paused access to its cloud backup system and is investigating the breach. As of Thursday, the company said it was working on restoring the service. ConnectWise disclosed the supply-chain compromise in an advisory on October 16 but had not initially identified who was responsible.
Analysis: This is the second AI-agent attack story on Chip Current this week — days ago, a single hacker reportedly let AI agents break into more than 100 company websites for about $5,000. The pattern Microsoft is describing now goes further: agents that are, in the company’s words, “not constrained by factors affecting humans” and can “work around the clock, on weekends, and on holidays and attempt multiple breach paths simultaneously.” When the offense runs at machine speed and the defense still staffs in shifts, the asymmetry — not the spyware — is the story.
