Fabraix has launched Nyx, a new autonomous red-teaming agent!
Users can point it at any AI agent and run adversarial attacks from a library of 10,000+ jailbreaks to surface security holes.
Every finding outputs the attack steps to the agent so it can spot a vulnerability. The same test can be replayed after a release to check if it was fixed.
Voice AI pentesting sample 🤖
we got an AI agent to exfiltrate customer bank balances and using an invoice that looked completely normal to a human. and we asked some people what they think.