Add the holistic threat model talk

README.md

@@ -165,6 +165,7 @@
 
 | Presentation Title | Author(s) | Year |
 | --- | --- | --- |
+| [Holistic ML Threat Models](presentations/Holistic%20ML%20Threat%20Models) | Adelin Travers | 2024 |
 | [Using Graph-Based Machine Learning Algorithms for Software Analysis](presentations/Using%20Graph-Based%20Machine%20Learning%20Algorithms%20for%20Software%20Analysis/) | Michael D. Brown                                     | 2023 |
 | [Exploiting Machine Learning Pickle Files](presentations/Never%20a%20Dill%20Moment:%20Exploiting%20Machine%20Learning%20Pickle%20Files) | Carson Harmon, Evan Sultanik, Jim Miller, Suha Hussain | 2021 |
 | [PrivacyRaven: Comprehensive Privacy Testing for Deep Learning](presentations/PrivacyRaven:%20Comprehensive%20Privacy%20Testing%20for%20Deep%20Learning) | Suha Hussain | 2020 |

presentations/Holistic ML Threat Models/README.md

@@ -0,0 +1,15 @@
+Holistic ML Threat Models
+=============================
+
+A common, but misguided threat modeling approach for systems incorporating ML models is to augment a typical system threat model with analysis of ML model-level attacks such as prompt injections. This effectively disconnects the ML model, the underlying system, and safety/privacy requirements in the threat modeling process and ultimately sets it up to fail by missing critical risks and controls to mitigate them. This talk presents Trail of Bits's approach to ML threat modeling and how it enables robust and practical control recommendations throughout the ML life cycle that account for the full complexity of the ML tech stack.
+
+Presented at:
+
+* [AI Village at BSidesSF, 2024](BSidesSF%20-%20AI%20Village%20-%20Holistic%20ML%20threat%20models.pdf)
+* [Graph the planet, 2024](Graph%20the%20planet%20-%20Holistic%20ML%20threat%20models.pdf)
+* [Tokyo AI (in Japanese), 2024](Tokyo%20AI%20event_%20総合的な機械学習の脅威モデリング.pdf)
+
+Authored by:
+
+* Adelin Travers
+