Learn how AI hackers detect fragility and how to thwart them with AI model resilience

Learn how AI hackers detect fragility and how to thwart them with AI model resilience On every lab test, your AI was superhuman. But how will it fare in the real world of smog, smears, and nation state hackers? In this session, we'll explore how AI hackers can measure the fragility of today's AI models, covering the model's vulnerability under real-world conditions across applications of varying data dimensions, from signals to images to videos. We’ll then show how to engineer robustness into models and sketch out tomorrow's AI supply chain where confidence is measurable and the model's perception can be inspected.