Vulnerability Research, Exploitation Capabilities of LLMs Still Lagging

Vulnerability research, exploitation capabilities of LLMs still lagging

Many large language models continue to lack vulnerability research and exploitation capabilities, with 48% and 55% of models failing the first and second VR tasks, respectively, while 66% and 93% failed the first and sec…

AI-driven cyberattacks remain rare as models show key flaws

AI-driven cyberattacks are currently rare due to key flaws in models, with research showing AI lacks reliability for complete autonomous hacks.