PNNL develops generative AI tool to outpace hackers

A group of researchers led by Pacific Northwest National Laboratory are utilizing generative artificial intelligence (AI) to protect from cyberattacks. The new tool accelerates the time it takes to set up defenses, according to an announcement written by Tom Rickey.

The tool, named Agentic LLMs for Offensive Heuristic Automation (ALOHA), was developed by PNNL researchers from large language model Claude through a partnership with Anthropic.

It works by “reconstructing complex cyberattacks” in hours. Rickey pointed out that hackers are also using AI for cyber-incidents. ALOHA analyzes cyberattacks and imitates them to “defend the system.”  Then, ALOHA “launches the rebuilt attack against the original target system in a siloed, offline environment to see if new protections that have been installed stop the attack.” This process usually takes several weeks. With ALOHA, it only takes a few hours, according to PNNL. The system also is more cost effective and reduces manual labor.

PNNL cybersecurity researcher Kristopher Willis told Rickey, “There are many programs out there to detect attacks. ALOHA goes much further, adapting attacks to particular hardware, software, and environments, giving guidance on the best way to protect those systems—and then attacking again and again to identify gaps in security and to enhance response.”