nullcon Goa 2017 - Training Machines For Application Scanning In Retro Style by Bharadwaj Machiraju

This talk explores the innovative use of machine learning in offensive security testing, addressing the limitations of current security tools that often lack application context. The speaker introduces both supervised and unsupervised machine learning techniques, illustrating their applications in developing a prototype application scanner that outperforms traditional crawlers. Attendees will learn about the scanner's ability to effectively spider applications, utilize feedback for fuzz testing based on learned knowledge, and the challenges faced during its construction. The session aims to demystify machine learning, empowering participants to contribute to or create their own intelligent scanning tools. Bharadwaj Machiraju, project leader for OWASP OWTF and an experienced web app security tool developer, presents his insights drawn from his extensive background in information security.