Brendan Dolan-Gavitt

Brendan Dolan-Gavitt is a distinguished engineer and AI researcher at XBOW, where he builds AI agents that find and fix software vulnerabilities before attackers can exploit them. He spent roughly a decade before that as an associate professor at NYU Tandon School of Engineering, researching software security and the security implications of AI-assisted coding tools such as GitHub Copilot. He holds a PhD from Georgia Tech and has previously worked at Columbia University, MIT Lincoln Laboratory and Microsoft Research. He co-authored StarCoder, a large language model for code generation, and helped build the AI system that became the first to reach number one on HackerOne's US leaderboard. On Secured by Galah Cyber, he discussed how AI penetration testing actually works and where it breaks down.