Google Threat Intelligence researchers have identified a hacker-operated AI framework, named exploit_agent, capable of discovering and exploiting software vulnerabilities without human intervention.
Google's Threat Intelligence Group expects a shift toward fully autonomous cyberattacks within six months as threat actors increasingly adopt agentic frameworks like ARTEX for large-scale intrusions.
A China-nexus threat actor has compromised 10 organisations globally using AI-enhanced autonomous tools, targeting the manufacturing, defence, and human rights sectors.
A China-linked threat actor has successfully compromised 10 organisations, including US defence contractors and semiconductor manufacturers, by deploying an autonomous AI penetration-testing framework to automate its operations.