Teaching AI to Find Real Vulnerabilities — David Brumley, Bugcrowd
David Brumley is using RL environments to teach AI to find vulnerabilities at machine speed
“we're pushing out programs faster than ever and so we need to be able to check them at machine speeds in scale”
CMU professor and Bugcrowd Chief AI & Science Officer David Brumley is applying two decades of cybersecurity research to designing reinforcement learning environments that teach AI to find real vulnerabilities. He draws a direct parallel between how elite human hackers learn—graduated difficulty, studying write-ups, iterative practice—and how RL agents should be trained. The talk signals growing serious academic and industry investment in autonomous vulnerability discovery, though the content is cut off before the technical methodology is fully presented.