China-Aligned Hackers Use AI Agents to Breach Thousands of Organizations via ConnectWise
Microsoft researchers report that China state-aligned hackers breached ConnectWise's cloud backup service and stole application credentials, then accessed roughly 2,250 customer organizations' Microsoft-hosted environments. The attackers used ARTEX, an open-source AI agent framework built for stress-testing defenses, to make nearly 3,000 sign-in attempts per minute, leaving defenders little time to respond. ConnectWise has paused its cloud backup service and is investigating.
Skynet Chance (+0.02%): An open-source agentic AI framework was repurposed for mass offensive cyberattacks by state actors. This shows autonomous AI systems can be weaponized at speeds that outpace human oversight, a marker of eroding control.
Skynet Date (+0 days): AI agents are already operating at machine speed in real-world cyber offense, suggesting autonomous-AI risks are arriving sooner than defensive capabilities can mature.
AGI Progress (0%): The attack demonstrates practical, scalable deployment of autonomous agents for complex multi-step tasks. However, it reflects existing capabilities rather than a fundamental advance toward general intelligence.
AGI Date (+0 days): Real-world agentic deployment confirms the current trajectory but does not meaningfully change the pace of AGI development.
<< All AI news for October 22, 2026
[ Get the daily index digest on Telegram → ]Related AI News
- OpenAI Launches Decisions API to Compete with TypeSafe's Jev in Fast, Cheap Decision-Model Market 2026-10-22
- Anthropic Launches Claude Duet, a Proactive Agent App for Professionals and Small Businesses 2026-10-22
- Meta Creates Non-Voting Stock Class to Fund Expansive AI Buildout 2026-10-21
- Patch the Planet Initiative Coordinates Global Response to AI-Agent-Discovered Open Source Vulnerabilities 2026-10-21
- OpenAI, PerturbAI, and Scripps Research Deploy Thousands of AI Agents to Mine Brain Cell Data 2026-10-21