What happened after 2k people tried to hack my AI assistant

A developer opened their AI assistant to 2,000 participants who attempted to hack it, documenting common attack vectors and jailbreak techniques. The experiment revealed the most persistent attempts and what vulnerabilities emerged.