Cybersecurity

Google Warnings: Fully Automated AI-Powered Cyberattacks on the Rise

Google's cybersecurity threat group revealed that malicious actors are adopting agentic-based offensive frameworks to automate vulnerability discovery and exploitation at ultra-high speeds.

2026-11-16
3 min read
8 views
certi.news Editorial Team
Google Warnings: Fully Automated AI-Powered Cyberattacks on the Rise

Google, represented by its Threat Intelligence Group (GTIG), warned that malicious actors are increasingly relying on agentic-based workflows and AI-powered offensive frameworks to automate vulnerability discovery and exploitation processes, accelerating attack speeds to machine speed.

Google reports indicated that the initial experimental use of AI in cyberattacks has shifted to widespread adoption by state-sponsored hacking groups and financially motivated organizations. Among the most prominent documented examples is the use of the open-source penetration testing framework called CAIRN in an attack targeting the US aviation sector, along with suspected involvement of North Korean-linked hackers in using the Claude Code tool to deploy malware.

Expectations of a Shift Toward Full Autonomy

Security research estimates that within the next three months, malicious actors will be capable of launching simple attacks requiring little to no human intervention, relying primarily on multi-agent AI systems. More dangerously, fully autonomous and advanced automated attacks are projected to emerge during the second half of this year.

This aligns with a previous warning issued by the Federal Bureau of Investigation (FBI) regarding multiple malicious actors weaponizing the ARTEX penetration testing framework since August, launching attacks targeting organizations in the government, defense, finance, technology, and healthcare sectors.

Exploitation of Open-Source Frameworks

Google explained that attackers are continually exploiting open-source penetration testing frameworks such as ARTEX, CAIRN, CyberStrikeAI, and HexStrike AI to execute adaptive attacks against internet-exposed endpoints vulnerable to low-complexity flaws, for which large language models can easily build exploits.

In a prominent example from Google's AI threat tracker report, hackers used a multi-agent system called exploit_agent to completely automate the process of searching for and exploiting vulnerabilities without human intervention against French websites operating in the media, entertainment, and education sectors.

What Does This Development Change Practically?

This technological shift places immense pressure on defense and cybersecurity teams, as the deployment of these offensive frameworks has dramatically reduced the time available for defenders to patch their systems. Furthermore, it generates false leads and fake information, such as what occurred with PHP2Shell vulnerabilities that produced hundreds of AI-generated proof-of-concept exploits, multiplying the burden of verification and hardening for system protection teams.

News source
BleepingComputer
Open original source ↗
c
Author

certi.news Editorial Team

In the same category

You may also like

View all news