I had a chat with Keith Hoodlet on #OpenSourceSecurity about a recent research paper titled "Why AI-generated vulnerability patches still require expert human review"
While AI can find security vulnerabilities at an alarming rate, it's not so great at fixing them
