AI Security is Not a Pipe(line) Dream | RSAC Virtual 2024
Scaling AI in the enterprise will require scaling how we secure them. CISOs are trying to apply current security frameworks to next generation AI products. This can only get us so far, because the models ship direct learnings from the original data.
Frederick will cover current and upcoming security frameworks designed specifically to help understand and control the risk of AI. He will discuss how we can move from “trust me” models to rigorous systems designed to constrain risk and provide evidence to demonstrate due care.
* Describe frameworks to help work with insecurity and compliance
* Describe similarities and differences between traditional software pipelines and AI
* Describe how different security tools and techniques fit into such models like federated learning and differential privacy.
