SOFTCODING

How to trick AI into being 99% confident in completely the wrong answer This is why "our model is accurate" isn't a security statement. Accuracy tells you how something behaves on data nobody tampered with. That's a different question to how it behaves on data someone built specifically to trick it.

Harriet Farlow is the CEO and Founder of Mileva Security Labs, Australia’s first AI security company

With a PhD in Machine Learning Security and her upcoming book, she is dedicated to building AI security into a field that can rise to one of the world’s greatest technical and geostrategic challenges.

Ms. Fateme Yoosefi | Ferdowsi State University

A unique view of the Avicenna Mausoleum, beautiful Hamedan

Pre-conference Reception OWASP® Foundation Global AppSec USA