AI agents have evolved from simple chatbots to complex entities that interact with various systems, calling APIs and accessing sensitive data. Traditional identity and access management (IAM) systems were built to handle human users, not these non-human actors. As a result, many of these systems struggle to adapt to the new demands that generative AI presents. The whitepaper highlights key areas where conventional identity frameworks fall short when managing GenAI. It emphasizes the need for enhanced security measures before, during, and after an AI agent logs in. Human-in-the-loop approvals and fine-grained authorization emerge as essential components to ensure security. These mechanisms help mitigate risks by requiring human oversight in critical decision-making processes. Auth0 positions itself as a solution for teams looking to secure AI workflows more efficiently. It offers tools that help implement the necessary security measures while maintaining flexibility in deployment. This approach aims to balance the growing need for security with the operational demands of rapidly evolving AI technologies.