theme image
Lessons Learned from Developing Secure AI Workflows at GoogleLessons Learned from Developing Secure AI Workflows at Google
  1. talk
  2. AI

Lessons Learned from Developing Secure AI Workflows at Google

Available Media

video

Slides (pdf)

Slides (Online)

ConferenceRSA - 2024
AuthorElie Bursztein

This talk discuss the Google Security AI Framework (SAIF), which is used to protect AI systems and workflows at Google. Through concrete examples, this presentation cover the key threats that must be addressed when deploying AI systems and what mitigations to put in place with a focus on when to extend existing systems versus adding AI specific protections.

newsletter signup slide

Get cutting edge research directly in your inbox.

newsletter signup slide

Get cutting edge research directly in your inbox.