Therefore, a more nuanced approach that considers both the technical and inferential privacy risks is necessary to ensure comprehensive data protection in the era of advanced AI.