/
← Voltar ao feed
medium financial

ThreatBook launches RedAgents AI pentest platform after Artex misuse in financial sector attacks

1 fonte(s) · 1 atualização(ões) · publicado 20/10/2026 03:20 · coletado há 4 dias

ThreatBook announced RedAgents on October 16, 2026. It is a SaaS platform that runs several AI penetration testing agents in parallel against authorized targets and shows the findings in one interface. The platform combines Artex Pro, Cyber(YanJi), Yanxun and CyberStrikeAI Enterprise Edition. ThreatBook says that combining agents improves coverage and reduces false positives, but the announcement gives no independent benchmark to support this. The launch has a dual-use backdrop: ThreatBook says the original Artex project became closed source on October 8 after attackers misused it in financial sector incidents, and separate reporting linked suspected abuse to attacks on South Korean financial institutions.

Detalhes técnicos

  • Artex Pro extends an autonomous penetration testing project developed by the RedAgents product lead.
  • Cyber(YanJi) is a Python-based agent that searches unknown state spaces for a route to a defined success condition. ThreatBook says it solved all 54 challenges in the Tencent Cloud Hackathon AI Penetration Testing Challenge.
  • Yanxun, from Changsha Yanzhi Technology, coordinates dozens of specialized agents for reconnaissance, code auditing, penetration testing and red team work. Its claims of reproducibility and zero false positives are vendor assertions.
  • CyberStrikeAI Enterprise Edition adds native MCP support and integrations with more than 100 security tools. ThreatBook acquired CyberStrikeAI in September 2026.
  • Stated controls include target allowlists that define testing boundaries, assessment data kept inside ThreatBook's environment, and Zero Data Retention support for complete deletion of test data. These are vendor claims and have not been independently verified.
  • Availability is limited testing, not general access.

Remediação

  • Treat AI-driven offensive tooling as a growing threat to exposed attack surface. Prioritize patching and reducing external exposure.
  • Financial institutions should monitor for automated, high-volume exploitation attempts.
  • Organizations adopting such platforms should check the allowlist and authorization controls and the data retention terms before use.

Linha do tempo

  1. gbhackers.com ↗ · pub. 20/10/2026 03:20 · coletado 20/10 05:00 primeiro registro

    ThreatBook launches RedAgents AI pentest platform after Artex misuse in financial sector attacks

    ler a análise desta fonte

    ThreatBook announced RedAgents on October 16, 2026. It is a SaaS platform that runs several AI penetration testing agents in parallel against authorized targets and shows the findings in one interface. The platform combines Artex Pro, Cyber(YanJi), Yanxun and CyberStrikeAI Enterprise Edition. ThreatBook says that combining agents improves coverage and reduces false positives, but the announcement gives no independent benchmark to support this. The launch has a dual-use backdrop: ThreatBook says the original Artex project became closed source on October 8 after attackers misused it in financial sector incidents, and separate reporting linked suspected abuse to attacks on South Korean financial institutions.

    Technical Details

    • Artex Pro extends an autonomous penetration testing project developed by the RedAgents product lead.
    • Cyber(YanJi) is a Python-based agent that searches unknown state spaces for a route to a defined success condition. ThreatBook says it solved all 54 challenges in the Tencent Cloud Hackathon AI Penetration Testing Challenge.
    • Yanxun, from Changsha Yanzhi Technology, coordinates dozens of specialized agents for reconnaissance, code auditing, penetration testing and red team work. Its claims of reproducibility and zero false positives are vendor assertions.
    • CyberStrikeAI Enterprise Edition adds native MCP support and integrations with more than 100 security tools. ThreatBook acquired CyberStrikeAI in September 2026.
    • Stated controls include target allowlists that define testing boundaries, assessment data kept inside ThreatBook's environment, and Zero Data Retention support for complete deletion of test data. These are vendor claims and have not been independently verified.
    • Availability is limited testing, not general access.

    Remediation

    • Treat AI-driven offensive tooling as a growing threat to exposed attack surface. Prioritize patching and reducing external exposure.
    • Financial institutions should monitor for automated, high-volume exploitation attempts.
    • Organizations adopting such platforms should check the allowlist and authorization controls and the data retention terms before use.