DKDKCISSP
AI SecurityINCIDENT

Google warns that fully-automated AI attacks are coming soon

Google warns that threat actors are increasingly using AI frameworks to automate attacks, leading to faster and more sophisticated exploitation of vulnerabilities.

DKCISSP News DeskBleepingComputer16 Nov 2026, 7:30 pm2 min read
Google warns that fully-automated AI attacks are coming soon
Image courtesy of BleepingComputer. Original report
BACKGROUND

Google's Threat Intelligence Group has reported a significant rise in the use of AI-based frameworks by threat actors to automate vulnerability discovery and exploitation, marking a shift toward fully automated attacks that require minimal human involvement.

WHAT WE KNOW

The rise of AI in cybersecurity threats has reached a new level, with attackers now employing agentic workflows powered by AI-based offensive frameworks. Google Threat Intelligence Group (GTIG) has observed that both financially motivated and state-sponsored groups are adopting these technologies on a large scale. This shift not only accelerates the speed of attacks but also enhances their efficiency, allowing for the automation of processes that were once labor-intensive.

A key example of this trend is the use of CAIRN, an open-source multi-agent penetration-testing framework, which has been linked to an attack against the U.S. aerospace sector. Additionally, suspected North Korean hackers have reportedly utilized Claude Code to deliver malware. Google predicts that within the next three months, threat actors will be able to initiate unsophisticated attacks with little to no human interaction, primarily relying on multiple AI agents. The expectation is that more sophisticated, fully autonomous attacks will likely emerge within the next six months.

In a recent advisory, the FBI highlighted that various threat actors have weaponized the ARTEX AI pentesting framework since August, targeting sectors such as finance, government, technology, defense, and healthcare. Google notes that attackers are systematically abusing open-source pentesting suites like ARTEX, CAIRN, CyberStrikeAI, and HexStrike AI to conduct fully automated attacks. These frameworks facilitate adaptive attacks against internet-exposed endpoints vulnerable to low-complexity flaws, which large language models (LLMs) can easily exploit.

GTIG's AI Threat Tracker report details how hackers recently employed a multi-agent system named 'exploit_agent' to automate vulnerability discovery and exploitation in attacks on French media, entertainment, and education websites. This marks the first time GTIG has identified threat actors using LLM frameworks to fully automate these processes without human oversight. The exploit_agent framework was discovered on an exposed staging server and included various malicious tools and stolen victim data.

Google warns that the deployment of such automated frameworks significantly reduces the time defenders have to patch vulnerabilities, while also generating false leads that complicate response efforts. An example cited is PHP2Shell, where hundreds of purported proof-of-concept exploits, mostly AI-generated, surfaced shortly after PHP announced two flaws. This situation places an added burden on defenders, who must verify the legitimacy of the code and develop appropriate protections.

WHAT WE DON'T KNOW YET
  • It is unclear how quickly organizations can adapt their defenses to counter these emerging threats.
  • The full extent of the impact of these automated attacks on various sectors remains unknown.
MORE IN AI SECURITY

More cybersecurity reporting

California sues OpenAI over Hugging Face attackBBC News · 17 Nov 2026, 4:42 amGemini AI agents repeatedly bypassed safeguards to attack real websiteBleepingComputer · 17 Nov 2026, 4:24 amGoogle Project Zero cuts bug disclosures to 30 days amid AI threatsBleepingComputer · 17 Nov 2026, 3:01 amCalifornia sues OpenAI over AI agents escaping sandboxes, hacking firmsBleepingComputer · 17 Nov 2026, 2:29 am