Hundreds of global institutions fell to an AI-powered cyberattack in under three hours

AI黑客帝国与全球防线的较量 https://t.co/ZArEXwSLfy
— 美国之音中文网 (@VOAChinese) October 22, 2026
Microsoft has warned that a hacking group linked to the Chinese government leveraged advanced artificial intelligence to breach defenses at ConnectWise, part of a sweeping automated cyberattack campaign targeting hundreds of organizations worldwide.
According to a detailed threat intelligence report released on Wednesday, October 21, by Microsoft’s security research division, the hacker collective tracked as Storm-0940 deployed an AI-powered penetration tool known as ARTEX throughout September. The group used the software to execute large-scale, automated assaults across multiple critical sectors globally.
Investigators found that many targeted organizations saw their cloud infrastructure thoroughly compromised, with attackers securing highest-level administrative privileges. Most alarming was the velocity of the breaches: moving from target identification to complete system takeover took less than three hours.
The targeted entities span a wide range of strategic industries, including aerospace and defense, finance, energy, automotive, government agencies, semiconductor firms, major technology companies, and prominent think tanks.
Further analysis indicates that Storm-0940 is directly connected to a recent supply chain security incident involving multi-tenant applications at ConnectWise, a major IT management and software provider.
ConnectWise CEO Manny Rivelo acknowledged the breach in a statement, noting that a group linked to the Chinese government likely targeted their cloud backup systems—a conclusion drawn from both Microsoft’s findings and the company’s internal evidence.
While the enterprise has disabled compromised application programming interfaces (APIs) and is working to restore operations, Rivelo conceded that emergency fixes have limitations. Remediation efforts cannot entirely eliminate illicit access paths already established within certain customer tenant systems, leaving hidden backdoors that must be hunted down and purged individually.
Microsoft’s report underscores the severity of the threat landscape. Data shows that while a majority of targeted institutions ultimately resisted full intrusion, Storm-0940 managed to pierce cloud defenses at numerous organizations in record time. Crucially, these acquired access credentials often retained their potency even after discovery and remediation.
ARTEX was originally engineered for network defense and security posture evaluation. However, its potency lies in its ability to integrate multiple advanced AI models to autonomously conduct reconnaissance, vulnerability hunting, and data exfiltration with minimal human intervention.
In response to media inquiries, a spokesperson for the Chinese Embassy in Washington reiterated Beijing’s consistent stance, stating that China firmly opposes and combats all forms of cyberattacks while rejecting the use of cybersecurity issues as a pretext to smear and discredit the country.
The Microsoft advisory followed an investigative report published by Reuters on Monday, October 19, which detailed a separate incident where a Chinese-speaking threat actor utilized AI agent technology to breach security defenses across more than 100 corporate websites with minimal manual oversight.
Snorre Fagerland, chief threat researcher at cybersecurity firm NorForge, observed that the attacker’s workflow was remarkably streamlined: by inputting brief prompts, the AI agent autonomously executed continuous vulnerability scans and launched attacks over a matter of hours—using instructions written entirely in Chinese that spanned fewer than 100 characters.
These developments highlight a sobering reality: the advent of advanced artificial intelligence threatens to exponentially amplify the destructive capacity of individual hackers, scaling their operational reach to rival historical cybercrime syndicates.
