Hackers weaponise ARTEX AI agents to automate network intrusions and data theft

Google Threat Intelligence Group◼ and Mandiant◼ report that threat actors are weaponising the ARTEX◆ autonomous penetration-testing framework to accelerate intrusions, privilege escalation and data theft. Post-compromise tasks are delegated to coordinated AI agents, compressing execution sequences into seconds. Activity has been observed since late August 2026. The source text is truncated and supplies no CVEs, affected versions, indicators of compromise or mitigations, so defenders should rely on the full GTIG and Mandiant reporting for detection guidance.

For security analysts · Written by AI from 1 source · Wrong?