Yves-Alexandre de Montjoye

Jan 9, 2025 | nature.com | Luc Rocher |Yves-Alexandre de Montjoye |Julien M. Hendrickx

AI techniques are increasingly being used to identify individuals both offline and online. However, quantifying their effectiveness at scale and, by extension, the risks they pose remains a significant challenge. Here, we propose a two-parameter Bayesian model for exact matching techniques and derive an analytical expression for correctness (κ), the fraction of people accurately identified in a population. We then generalize the model to forecast how κ scales from small-scale experiments to the real world, for exact, sparse, and machine learning-based robust identification techniques. Despite having only two degrees of freedom, our method closely fits 476 correctness curves and strongly outperforms curve-fitting methods and entropy-based rules of thumb. Our work provides a principled framework for forecasting the privacy risks posed by identification techniques, while also supporting independent accountability efforts for AI-based biometric systems. Advanced machine learning techniques have demonstrated the identifiability of human traces online, however, assessment of their potential risks is usually done with small-scale datasets. The authors propose a physics-based approach to evaluate the effectiveness of identification techniques from reported measurements.