AntiFox AI Pentesting Platform Used for AWS Credential Theft and Lateral Movement
Threat actors are weaponizing the AntiFox agentic AI platform to automate the exploitation of internet-facing IIS servers, facilitate web shell deployment, and conduct lateral movement into AWS environments via stolen credentials.
Triage facts
Revision 1- Applicability
- Internet Information Services, AWS
- Evidence
- Unknown
- Severity
- High
- Action
- Immediate escalation
- Sources
- 1 public source
- Freshness
- Revision 1Nov 11, 16:44 UTC
- Source quality
- B2Very likely
Arctic Wolf Labs has identified a malicious campaign leveraging AntiFox, an open-source AI-driven autonomous penetration testing platform. The attackers use AntiFox to automate post-exploitation tasks, including configuration discovery, AWS service enumeration, and secret exfiltration. The platform, available on GitHub, enables threat actors to execute multi-stage engagement workflows with minimal human intervention.
Attack Chain
- Initial access is achieved via an undisclosed vulnerability in an internet-facing IIS application.
- The actor deploys an ASPX web shell using the w3wp.exe IIS worker process.
- The agent uses the web shell to execute operating system commands to locate web.config files and configuration backups.
- Sensitive data, including application settings and database connection strings, are read and exfiltrated to a web-accessible directory.
- The actor attempts to connect to internal production databases using credentials extracted from the application configuration.
- The actor leverages stolen or discovered AWS IAM identities to enumerate cloud services (e.g., S3, Secrets Manager, RDS) and exfiltrate production secrets via API calls.
Impact
Observed impact includes the compromise of internet-facing IIS web servers, unauthorized access to application configuration files and database credentials, and confirmed exfiltration of production secrets from AWS environments. Attackers successfully leveraged AWS API calls for service enumeration and secret extraction.
Recommendation
- Audit all internet-facing IIS servers for unauthorized ASPX files in web-accessible directories, particularly those created by the w3wp.exe process.
- Implement strict egress filtering on web servers to prevent unauthorized internal database connections and AWS API service communication.
- Review AWS IAM logs for anomalous enumeration activity (e.g., Describe*, List* operations across multiple services) followed by unauthorized GetSecretValue requests.
- Monitor GitHub repositories and developer tools for the presence of autonomous penetration testing frameworks like AntiFox, which may be repurposed by malicious actors within your environment.
Immediate actions
Scan web-accessible directories on all internet-facing IIS servers for unauthorized ASPX files.
Audit CloudTrail logs for unusual API call volume, specifically focusing on GetSecretValue and service enumeration patterns.
Threat Hunt
Identify w3wp.exe spawning shell processes (cmd.exe, powershell.exe) or creating new files in the web root.
Data: Process creation telemetry from IIS hosts, File system auditing on web root directories