Security for AI (artificial intelligence) is a major concern for companies that are increasingly using AI-based systems and devices. As AI becomes more integrated into our lives, all stakeholders, from developers to end users, share a responsibility. AI systems face a variety of risks, ranging from direct attacks such as model theft to stealth threats such as model deception or data poisoning.

In this talk, we will look at the threats to AI/ML models and the real-world consequences of insecure AI systems, and present effective measures needed to assess, detect and protect against these threats. We will also explore the importance of a security-by-design approach and the adoption of security standards and best practices for achieving AI security at scale. 

...