Google Finds Hackers Using AI Agents to Build and Run Exploits From Software Patches

Ravie LakshmananNov 16, 2026Agentic AI / Cyber Attacks

Google on Monday revealed that adversarial use of artificial intelligence (AI) by threat actors has gone beyond manual prompting to deploying autonomous multi-agent offensive frameworks like A

Google Finds Hackers Using AI Agents to Build and Run Exploits From Software Patches

Google Finds Hackers Using AI Agents to Build and Run Exploits From Software Patches

Ravie LakshmananNov 16, 2026Agentic AI / Cyber Attacks

Google Finds Hackers Using AI Agents to Build and Run Exploits From Software Patches

Google on Monday revealed that adversarial use of artificial intelligence (AI) by threat actors has gone beyond manual prompting to deploying autonomous multi-agent offensive frameworks like ARTEX, CAIRN, CyberStrikeAI, and PentAGI to execute multi-stage intrusions at greater operational velocity.

In tandem, threat actors are leveraging the power and capabilities of large language models (LLMs) and agentic frameworks to automate the reverse engineering of patches via differential analysis, allowing them to rapidly weaponize newly disclosed vulnerabilities, as observed in the case of PaperCut NG/MF flaws earlier this September.

The abuse of open source, agentic penetration-testing (pentesting) frameworks is seen as a way to “automate discrete phases and specific operational tasks during live intrusions,” Google Threat Intelligence Group (GTIG) said in a report shared with The Hacker News.

The tech giant said the adoption represents a paradigm shift in which cyber attacks take place at machine speed, allowing the operations to be scaled significantly by chaining together varied attack paths even if the underlying mechanics remain the same.

A number of threat actors have been observed experimenting with these frameworks and harnesses, including financially motivated and suspected espionage actors aligned with China.

“These observations additionally extend to the use of other tools and services, such as ZeroHollow, an inference service used for exploit development and LLM routing, and Agent Zero, an open source agent framework which features autonomous code execution and dynamic tool generation,” Google said.

Another key area of misuse revolves around automating labor-intensive tasks like vulnerability analysis and proof-of-concept (POC) weaponization using LLM-driven pipelines, allowing threat actors to collapse the window between patch disclosure and active exploitation.

“Threat actors may find it more efficient to use LLMs to rapidly weaponize patched n-day vulnerabilities, as they can focus research on a finite set of source materials: comparing source code versions (including from pre-release or preview channels), and examining release notes, security research, and POC code related to specific flaws,” Google added.

Simultaneously, threat actors are also targeting AI infrastructure to illicit profit from LLMjacking attacks by stealing LLM credentials and cloud resources from compromised enterprise environments, and then reselling the access through underground proxy services like APIMart and JetNight at discounted prices, otherwise known as LLM mixers.

The findings show that the emergence of open-source frameworks have further lowered the barrier to entry, enabling bad actors to easily integrate new exploits, autonomously scale attacks, and mount convincing phishing campaigns, further compressing attack timelines.

This includes a Chinese-language phishing platform called Jieyou that has integrated an LLM chatbot assistant to allow customers to configure phishing panels and modify existing templates using natural language commands without necessarily having the technical expertise to do so.

“The evolution of the Jieyou platform highlights a growing industry trend where PhaaS providers leverage LLMs as a modular force multiplier,” Google said. “Rather than manually editing payloads, low-skilled affiliates can now utilize natural language through the chatbot to dynamically customize pretext lures and phishing interfaces on the fly.”

About Author

What do you feel about this?

Subscribe To InfoSec Today News

You have successfully subscribed to the newsletter

There was an error while trying to send your request. Please try again.

World Wide Crypto will use the information you provide on this form to be in touch with you and to provide updates and marketing.