Privacy perception plays a pivotal role in shaping user behaviour and trust in digital environments. This work investigates how eye movement data can be used to infer users’ subjective sense of privacy. We propose and evaluate a set of tasks that explore key dimensions of perceived privacy, including detection, quantification, contextual influence, private attribute recognition and Perceived Privacy Knowledge. By leveraging gaze as a dynamic and implicit signal, our approach offers a novel pathway for understanding privacy preferences without relying on self-reported data. The findings highlight gazebased modelling as a promising direction for developing privacy-aware systems that are more responsive to individual concerns and contexts.