Hackers abused AntiFox AI pentesting tool to steal credentials and access AWS environments, Arctic Wolf said
Security researchers Arctic Wolf said unnamed threat actors ran their own instances of AntiFox, an open source AI-powered penetration testing platform, on Alibaba Cloud infrastructure. They used it to analyse compromised systems, search for passwords and access victims' AWS cloud resources. In some attacks they first compromised internet-facing Microsoft IIS web servers and installed web shells; in others they used AWS credentials obtained elsewhere. Amazon said AWS services were not compromised and that the attackers used credentials taken from customers' compromised applications.
What the AI did
Attackers ran their own copies of AntiFox, an open source AI tool built for testing how easily systems can be broken into, on Alibaba Cloud servers. They used it to analyse systems they had already broken into, search for passwords and reach victims' Amazon Web Services (AWS) cloud resources.
First reported November 12, 2026 · Added to the register November 17, 2026 · 1 source
- Developer
- Unnamed Chinese developer
- Model
- AntiFox
- Affected
- Internet-facing Microsoft IIS web servers
- When it happened
- Not stated in the sources
- First reported
- November 12, 2026
What this means for you
Could this affect you?
If you run internet-facing Microsoft IIS web servers or keep AWS credentials in application settings. Organisations running Microsoft IIS web servers reachable from the internet, or storing AWS credentials in their applications, were exposed, and their databases, storage and other cloud resources may have been accessed.
What to check
- Monitor web servers for web shells, which are hidden scripts that give attackers remote control, and for unapproved system commands.
- Investigate unusual activity in your AWS accounts.
- Rotate any exposed passwords, API keys and cloud credentials.
- Limit each account to the permissions it needs and turn on detailed logging.
Areas of your AI programme this touches
Every fact and its source (4)
- Tool originFork of ARTEX 0.1.1“AntiFox was built as a fork of ARTEX 0.1.1”[1]
- Source of findingsArctic Wolf researchers“This is according to security researchers Arctic Wolf, who recently disclosed their findings on GitHub”[1]
- Amazon responseAWS said its services were not compromised“AWS services were not compromised and operated as designed.”[1]
- Attacker infrastructureSelf-hosted on Alibaba Cloud“threat actors were hosting their own instances of the tool on Alibaba Cloud infrastructure”[1]
Sources
- AntiFox AI agents hijacked by hackers to carry out theft of AWS credentialstechradar.com · November 12, 2026
How this record is classified. Severity S (4/5): real harm to one organisation, data exposed, or a containment failure. OECD level: incident, an event in which an AI system led to actual harm. Evidence: Reported, meaning press reporting only.
The facts in At a glance and in Every fact and its source are quoted from the sources listed. The summary, What the AI did and What this means for you are written from those sources and checked against them automatically before publication. Records reflect what has been disclosed, not everything that has happened. Incidents are found through news feeds and the GDELT Project.

