We've spent years solving the problem of verifying who is accessing a system. But what happens when the entity requesting access can fabricate its own context, misrepresent its purpose, and construct a convincing identity that no MFA prompt will ever catch?

AI agents don't just authenticate. They reason, communicate, and act. This creates an entirely new class of identity risk that the industry has barely begun to address: agents that deceive. Whether through adversarial manipulation or other means, AI agents can present false claims about their identity, their delegated authority, and their intended actions–all while passing every traditional security check.

In this session, we'll unpack the emerging threat landscape around AI agent intent, explore why existing controls fall short. The identity perimeter has moved. This session will show you where it's going, and what to build next.