I ASKED 100 AGENTS TO HACK ME

Summary from the TLDR AI newsletter

Around 100 self-hosted agents attempted to hack various online accounts over five hours. They compromised three accounts through software vulnerabilities and two through password brute-forcing, while also making 16 social engineering attempts. The experiment tested abliterated open-source models' capabilities, revealing potential future risks as these models improve and become cheaper to deploy.