ThreatBook Launches RedAgents Cloud Platform for Multi-Agent AI Penetration Testing
ThreatBook introduces a new cloud service that orchestrates multiple AI hacking agents to uncover vulnerabilities and test attack paths in authorized environments.
Key points
- RedAgents announced October 16, 2026, unifies four distinct AI penetration testing tools into one cloud-based service for security teams.
- The platform aims to solve incomplete visibility by running diverse agents in parallel to identify weaknesses a single engine might miss.
- Included agents are Artex Pro, Cyber(YanJi), Yanxun, and CyberStrikeAI Enterprise Edition, offering varied search strategies and operational strengths.
ThreatBook announced RedAgents on October 16, 2026, introducing a cloud-based platform that deploys multiple artificial intelligence agents for authorized penetration testing. The service allows security operations teams to run several distinct AI agents simultaneously against specific targets to discover vulnerabilities and map potential attack paths. This launch addresses the growing need for comprehensive automated security assessments within enterprise environments.
The platform integrates four specific AI agents: Artex Pro, Cyber(YanJi), Yanxun, and CyberStrikeAI Enterprise Edition. According to GBHackers, this consolidation provides a single environment for security teams to manage complex testing engagements. Users can deploy agents individually, select specific combinations for particular assessments, or run all agents in parallel against the same target. This flexibility aims to reduce the administrative burden of managing separate tools and vendor integrations.
How the Multi-Agent Approach Works
ThreatBook founder and CEO Feng Xue stated that no single agent can discover every security issue. The company argues that combining agents with different search strategies improves overall vulnerability coverage while reducing false positive results. By examining the same attack surface through multiple operational lenses, the platform seeks to identify attack paths that individual agents might overlook.
The technical proposition relies on diversity rather than a single unified model. Agents with different strengths and methodologies analyze targets concurrently. Vulnerability findings, execution logs, and assessment data appear in a unified interface, eliminating the need for teams to manually consolidate results across disparate platforms. This approach supports more frequent assessments as AI penetration testing capabilities continue to evolve.
Included AI Agents and Capabilities
Artex Pro extends an autonomous penetration testing project developed by the RedAgents product lead, focusing on independent vulnerability discovery. CyberStrikeAI Enterprise Edition emphasizes execution control, operational stability, and traceable results alongside its autonomous discovery features. These agents provide a foundation for structured and reliable testing processes within the platform.
Cyber(YanJi) is a Python-based agent designed to search unknown state spaces for viable routes from an initial position to a defined success condition. ThreatBook reports that this agent solved all 54 challenges in the Tencent Cloud Hackathon AI Penetration Testing Challenge, ranking first among individual participants. This performance highlights its capability in complex, structured testing environments.
Yanxun, developed by Changsha Yanzhi Technology, targets extended penetration scenarios. Its development is associated with research from Peking University’s Changsha institute. This agent adds specialized capabilities for deeper, more prolonged assessment workflows. Together, these tools form the core of the RedAgents orchestration layer, aiming to provide more complete security visibility than single-engine solutions.
As Threat Times reported on October 19, 2026, other IT services firms are also expanding AI agent deployment for enterprise production use, indicating a broader industry shift toward operational AI. RedAgents focuses specifically on the offensive security testing domain, providing a specialized platform for vulnerability discovery.
What should you do?
Security professionals should evaluate whether current automated testing tools provide sufficient coverage for their unique attack surfaces. Organizations using AI-driven security assessments should monitor for new platforms that offer multi-agent orchestration capabilities. Teams should ensure all penetration testing activities are strictly authorized and scoped appropriately before deployment.
What to do and how to stay safe: ThreatBook
- Review current automated vulnerability scanning tools to identify gaps in coverage that single-engine solutions might miss.
- Assess the need for parallel agent deployment in your security testing strategy to improve detection of complex attack paths.
- Ensure all AI penetration testing activities are conducted within strictly defined, authorized scopes to prevent unauthorized access.
- Monitor vendor announcements for updates on multi-agent platforms and their integration capabilities with existing security operations centers.
Step-by-step guide: Risk-Based Vulnerability Management: Prioritize Fixes by Actual Danger
General security guidance from the Threat Times newsroom. It is not confirmed advice from the organisations named in this story.
Frequently asked questions
What is RedAgents?
RedAgents is a cloud-based platform announced by ThreatBook that deploys multiple AI penetration testing agents against authorized targets to discover vulnerabilities.
Which AI agents are included in RedAgents?
The platform includes Artex Pro, Cyber(YanJi), Yanxun, and CyberStrikeAI Enterprise Edition, each offering different search strategies and operational strengths.
Does RedAgents guarantee better detection rates?
ThreatBook argues that combining agents improves coverage and reduces false positives, but no independent comparative benchmark establishing aggregate detection improvement has been provided.
