PLOT4AI 2.0: Open source Threat Modeling for Trustworthy AI (WHY2025)

PLOT4AI 2.0 is a pioneering open source AI threat modeling tool that provides a structured, lifecycle-based approach to AI risk identification. With over 100 AI-specific risk sources across eight categories, it aligns with the EU AI Act and supports trustworthy AI development and deployment. In this talk, the author will present the story of this internationally recognized tool, first published in 2022, and will introduce its new, expanded 2.0 version. More info @ https://plot4.ai/ After three years of research, in 2022 the first version of PLOT4AI launched with 86 AI-related threats. At that time AI security was still a niche topic discussed mainly by a few and AI safety was barely recognized beyond robotics and reinforcement learning. Then, just seven months later, ChatGPT launched, and the AI landscape changed overnight. Suddenly, AI became a central topic in public discourse, governance, and policy. The EU AI Act entered the scene, putting fundamental rights at the heart of AI product regulation. What was once a niche technical concern had become a global geopolitical issue, influencing regulatory and economic agendas around the world. It became clear: PLOT4AI needed a major update. In this talk, the author of PLOT4AI will take you behind the scenes of the tool’s creation and introduce PLOT4AI 2.0: a major new release of this open source AI threat modeling framework. The updated version includes over 138 AI-related threats, including threats related to Generative AI, Agentic AI, and complex deployment environments. PLOT4AI isn’t just a tool, it’s a collaborative effort to make AI safer for everyone! As an open source initiative, it's built on feedback, shared experience, and contributions. Whether you’ve spotted a missing threat, devised a new mitigation, or have real-world examples to add, your input is welcome and encouraged! This talk is both a deep dive into the evolution of AI threat modeling and a call to action for the AI open source communities to shape safer, more accountable AI together. Licensed to the public under https://creativecommons.org/licenses/by/4.0/ about this event: https://program.why2025.org/why2025/talk/KB7ATS/