Skip to main content
Technical mechanism diagram: the Cairn AI-agent attack path from exposed instance to second university breach. Illustration: SecurityXP

Chinese-Nexus Hackers Used an Exposed Cairn AI Instance to Find and Exploit an Apparent Zero-Day

· 1 min read · SecurityXP

Key Facts

  • An agent that discovers and exploits a zero-day in minutes “completes its activity faster than many security operations centers can manually triage an alert,” Moses warned, noting that the agents “ran continuously for 16 hours without breaks, sleep, or working-hour constraints.”
  • On September 10, 2026, Amazon Threat Intelligence identified a server with a publicly hosted instance of Cairn, an open-source autonomous AI penetration-testing framework, running without authentication.
  • “Amazon Threat Intelligence assesses with high confidence that it was a zero-day based on the absence of any corresponding fix, issue, advisory, or prior mention in vulnerability databases for the affected open-source project.” The attacker then used the same apparent zero-day to breach a second university.
  • “At the time of discovery, this path traversal vulnerability in the medical imaging application did not correspond to any assigned security advisory and was unpatched,” Moses said.
  • The only successful attempt occurred when the agents discovered a vulnerability in an open-source medical imaging AI application and exploited it.

Amazon Threat Intelligence says Chinese-nexus hackers used Cairn, an open-source autonomous AI penetration-testing framework, to discover an apparent zero-day while breaching several universities in Asia. On September 10, 2026, the team identified a server hosting a public instance of Cairn that was running without authentication.

“Obtain shell access”

The operator handed the agents a single goal: “obtain shell access.” Moses said the agents “ran continuously for 16 hours without breaks, sleep, or working-hour constraints,” and that an agent able to find and exploit a zero-day in minutes “completes its activity faster than many security operations centers can manually triage an alert.”

The only successful attempt came against an open-source medical imaging AI application, where the agents found a path traversal vulnerability and used it to gain full control of a server. The flaw “did not correspond to any assigned security advisory and was unpatched” at the time of discovery, Moses said.

High-confidence zero-day

Amazon assesses with high confidence that the flaw was a zero-day “based on the absence of any corresponding fix, issue, advisory, or prior mention in vulnerability databases for the affected open-source project.” The attacker then reused the same path traversal to breach a second university.

The AWS report cites a September study by Gambit Security that documented a financially motivated group using Cairn to breach online retailers. Researchers and law enforcement have issued multiple warnings in recent weeks about AI-backed offensive tools, particularly those from China, a category that has also included malicious agent-framework packages on PyPI.

Tags: Cairn AI agents zero-day Amazon Threat Intelligence China-nexus higher education

Sources & References

S SecurityXP
SecurityXP Cybersecurity News & Analysis

SecurityXP delivers cybersecurity news, vulnerability analysis, data breach reports, and threat intelligence.

Security Digest

The most important cybersecurity stories, CVE alerts, and threat intelligence — one concise email. Unsubscribe anytime.

Related Articles