AI has sharply increased the complexity and workload facing data security teams. Data moves faster, access is broader, and AI agents are entering environments with identities of their own, multiplying both exposure and noise.

What’s holding them back isn’t effort or expertise. As one security leader put it: “We can’t solve this by throwing bodies at it. We need the truly sensitive things to float to the top so we can act.” 

Instead, highly skilled teams burn cycles chasing false positives, tuning policies, and digging through low-severity alerts. More time is lost implementing those policies across different tools for different channels – email, endpoint, and cloud – each with its own workflow.

This is a practical discussion for leaders / practitioners on why that basic question has become so hard to answer, and how teams are changing their DLP operating model to surface real risk faster – without burning out the people running it.