Lucija Valentic (ReversingLabs): The Hidden Threat: Unmasking Malware in Machine Learning Models

Behind every shiny new machine learning model, there is code and many technologies, each with its distinct vulnerabilities and potential security pitfalls. It was just a matter of time before machine learning models would become a new attack surface for threat actors. If you'd like to learn more, join us in a session with Lucija Valentić, Threat Researcher from ReveringLabs. In this session, we'll go through a brief overview of ML formats, and demonstrate how and why some of the most common ones are unsafe and prone to security compromises. We'll also present solutions to this problem and cover what formats to use instead and why, which tools can help, and what general good development practices are for ML engineers and data scientists from the cybersecurity perspective.