Breaking

AI agents compromised 440 PaperCut instances across 395 organizations

September 10, 2026 at 11:57 EDT

Security
AI Agents

A likely Russian-speaking threat actor used hundreds of AI agents to exploit two PaperCut NG/MF vulnerabilities, compromising at least 440 instances hosted by 395 identified organizations in 48 countries, according to the September 9 analysis. The campaign shows how agentic tooling can compress exploit development and intrusion operations, while its uneven results expose the limits of autonomous attacks.

September 10, 2026 · PaperCut NG/MF

AI agents targeted PaperCut worldwide

Hundreds of autonomous agents compromised at least 440 PaperCut instances across 395 identified organizations in 48 countries.

The campaign reached far beyond education

Compromised PaperCut instances, shown with proportional vertical columns.

440

all compromised instances

204

education instances

Education accounted for 204 of 440 instances.

440

compromised instances

395

identified organizations

countries reached

26 sec

11 organizations compromised

What still worked

Traditional controls held. Cloudflare’s WAF blocked at least one exploitation attempt, and endpoint protection stopped further execution in some cases.

Where risk remained

Automation was uneven. Only 12 victim organizations reached domain administrator access, and the outcome varied by environment.

Continue reading

The rest of this article is for AI News Blitz readers. Choose an option below to keep reading.