VMTech
Discuss a project →

AntiFox-Linked Intrusions Access AWS Production Secrets

AntiFox-Linked Intrusions Access AWS Production Secrets

Arctic Wolf Labs has reported intrusions associated with AntiFox, an AI-driven autonomous penetration-testing platform, in which attackers accessed Amazon Web Services credentials, moved laterally through affected environments and, in some cases, made successful GetSecretValue requests against production secrets. The activity involved Alibaba Cloud systems hosting AntiFox interfaces, compromised internet-facing web services and authenticated AWS API activity.

AntiFox is available on GitHub and describes itself as a multi-agent platform built around a planner, workers and a main agent. Its stated workflow spans reconnaissance, attack-surface discovery, exploitation, evidence capture and report delivery. At the time of reporting, the project had 251 forks and 474 stars.

IIS compromise led to configuration discovery

In some observed intrusions, initial access was associated with a compromised internet-facing Microsoft Internet Information Services application. Arctic Wolf said the attackers used w3wp.exe, the IIS worker process, to create an ASPX web shell and execute operating-system commands. The precise vulnerability that enabled code execution on the IIS server is not known.

Commands executed through the worker process identified the execution context with whoami, enumerated application directories, checked permissions with icacls and tested whether files could be written to the web root. PowerShell commands searched for web.config files and backups, while the operators read application settings and database connection strings, enumerated debug and error-log directories, and copied configuration and log files into a web-accessible upload directory.

Arctic Wolf also observed an attempt to connect to an internal production database using credentials extracted from application configuration. The creation of the web shell and subsequent command execution through the same IIS worker process indicated that the affected web application had been compromised.

Valid IAM identities enabled cloud enumeration

Other activity began with valid AWS credentials obtained by an undetermined method. The identities were used to validate credentials and enumerate services including IAM, EC2, S3, Lambda, Secrets Manager, KMS, RDS, ECS, EKS, Redshift, Athena, Glue, Systems Manager, API Gateway, CloudWatch, CloudFormation, DynamoDB and Amazon Bedrock.

Most of the authenticated cross-service API requests returned AccessDenied. However, Arctic Wolf said some discovery and secret-access operations succeeded. In at least some cases, the activity went beyond Secrets Manager enumeration to a successful request for a production secret. A valid AWS identity was also used for S3 discovery and object operations through API calls.

Agentic tools change the operational tempo

Arctic Wolf said agentic frameworks can make penetration-testing techniques more accessible to threat actors and increase the speed and consistency with which they are applied in live attacks. AntiFox is part of a broader group of open-source agentic attack frameworks that connect language models to tool loops capable of acting on outputs autonomously.

The practical implication for businesses is to assess internet-facing applications, configuration-file exposure and AWS identity permissions as a connected path: investigate IIS web-shell indicators, limit access to production secrets, and review authenticated cloud API activity for unexpected enumeration and object operations.

#cloudsecurity#awssecurity#agenticai#websecurity
Open analytics
On the site 3 views
min read 4 11.11.2026
Instagram

AntiFox-Linked Intrusions Access AWS Production Secrets

Open the post on Instagram ↗