Bringing just-in-time authorization together with intent classification, so an AI agent's access decisions account for what it's trying to do, not just who it is.
Authorization without intent is a guess
Identity and intent are two halves of the same picture when it comes to agents. Knowing who's making a request only tells you half the story; the other half is whether the request itself makes sense for the task the agent was given. Ping knows the identity: who is trying to access what, and whether they're allowed to do so. Proofpoint knows the intent: what the request is actually for, and whether it matches what the agent was asked to do. Together, the two complete the picture. A single employee might operate just one identity, but that same employee can stand up many agents, each autonomously retrieving data, calling APIs, and taking action on systems inside the enterprise. More agents means more requests, and more requests means more chances for one of them to act outside what it was actually asked to do.
Proofpoint and Ping Identity already help joint customers act on risk at the moment of authentication. Now, we're extending that relationship into the agentic era, bringing Ping's authorization intelligence together with Proofpoint's intent classification, so authorization decisions factor in identity and intent together.
What we're building together
When an AI agent requests access to a resource, Ping's authorization layer would incorporate a verdict from Proofpoint on whether the request's intent matches the agent's assigned task. Ping would then use that verdict, alongside its own authorization checks, to approve or deny the request in real time, at the moment access is requested rather than after the fact.
The goal is to close the gap between authorization and intent without adding friction: instead of a separate review after access has already been granted, the decision happens once, informed by identity and intent together. We'll share more detail on the integration as it takes shape.
Looking ahead
For most organizations, the hard part of securing AI agents isn't standing up a new authorization system. It's making the one you already trust smart enough to account for intent, not just identity. We expect to share more on this integration in early 2027. If you're a joint Proofpoint and Ping Identity customer, reach out to your account team to learn more.