Attribute Inference

Attribute inference focuses on predicting sensitive personal attributes from readily available data, posing significant privacy concerns in the age of readily accessible large language models and readily available data. Current research investigates the effectiveness of various machine learning models, including graph neural networks and other deep learning architectures, in performing these inferences, as well as developing methods to mitigate these risks, such as data synthesis and pre-processing techniques aimed at preventing sensitive attribute prediction. This field is crucial for understanding and mitigating the privacy implications of increasingly powerful AI systems and for developing fairer and more privacy-preserving data analysis techniques.