To maintain trust, compliance, and accurate data exchange across the Web, it is essential that humans, organizations, and autonomous agents can correctly interpret and act upon complex policies. Without interoperable policies that are both human- and machine-readable, particularly in the exchange of personal data, there is a significant risk of misinterpretation, conflicting actions by autonomous agents, and legal uncertainty—threatening the reliability, security, and scalability of decentralized ecosystems. Addressing these issues requires policy-aware agents capable of supporting scalable and (semi-)automated negotiation of data exchange terms on the Web. Core challenges in this area can be grouped into four key functions: compliance checking, consistency checking, requirements checking, and negotiation. These capabilities are essential for agents to understand, reason over, and act on data usage policies. Among these, requirements checking plays a particularly crucial role, as it involves querying policy sets to identify granted or denied actions, obligations to fulfil, and applicable constraints—ultimately helping make policy decisions transparent and explainable to both humans and machines.