Security Foundation: How to secure your agents with an identity-first approach? - Seguridad Mania.com - España y América Latina

The massive growth of non-human/agentic identities creates an exploding attack surface, with most identities over-permissioned. The rise of agentic AI amplifies this risk, introducing vulnerabilities like "confused deputy" and prompt injection that expose privileged data. Securing the entire AI supply chain demands a shift away from broad, static service accounts. This webinar details how to protect your AI innovation with Google Cloud's comprehensive, "identity-first" security stack. We'll explore a platform based on robust IAM and resource management. We will show you how to secure the entire AI lifecycle without static keys using Workload Identity for a strong, attestable identity for every component. To counter agent intent risks, we introduce a vision of ephemeral, just-in-time identities. These capabilities, combined with Gemini-powered tools, help maintain a strong security posture, improve permissions hygiene, and build defense-in-depth for your AI workloads. What you will learn: > How to manage the explosion of non-human identities with a modernized, identity-first security stack and federated authentication. > The critical role of Universal Workload Identity in creating a verifiable chain of trust without static keys. > Strategies to mitigate novel AI risks like the "confused deputy" by moving towards ephemeral, just-in-time identities that secure an agent's intent. > How to build defense-in-depth using a full suite of access management controls (IAM Allow/Deny, PAM, Principal Access Boundary). Join us to learn how to leverage Google Cloud IAM to secure your entire AI ecosystem and innovate with confidence.