AI Threat Modeling With PLOT4ai: A Holistic Approach to Identifying Risks in AI Systems

June 29, 2024
1:35 PM
Stage 1
Hands-on workshop

About this session

Join this interactive workshop and learn about AI threat modeling using PLOT4ai (https://plot4.ai) and how OWASP Threat Dragon can be used alongside PLOT4ai. Through a hands-on exercise, you'll identify risks during the design phase of an AI system in real-life use cases. Collaborating in groups, you'll apply your newfound knowledge using the card game of PLOT4ai. You'll come away with a solid understanding of AI risks through a holistic approach and recognize the role that threat modeling plays in the AI landscape.

*What's PLOT4ai?*  PLOT4ai, an acronym for Privacy Library of Threats for Artificial Intelligence, is a pioneering threat modeling methodology inspired by LINDDUN and specifically designed to facilitate the creation of Responsible AI systems. This methodology encompasses 86 distinct AI threats, organized into 8 categories: Technique & Processes, Accessibility, Identifiability & Linkability, Unawareness, Non-compliance, Ethics & Human Rights, Safety, and Security, including specialized threats for generative AI and procurement of third-party AI systems.

About the speaker

About the speakers

Isabel Barberá is the author of PLOT4ai. She is also one of the members of ENISA Working Group on Data Protection Engineering and a National Expert at ISO/CEN/CENELEC working on the development of standards related to AI and privacy engineering.

Speaker

Speakers

Isabel Barberá
AI Advisor, Privacy Engineer & Co-founder, Rhite