#1761897: Webinar | Many Paths, One Policy: Enforcing AI Agent Actions at Runtime
| Description: |
AI agents are already taking actions inside business applications using legitimate access, often faster than security teams can review. As adoption accelerates, leaders face a critical challenge: how to govern what agents do without becoming a barrier to innovation. The problem is that security controls don't always follow agents across tools. Gateway-only policies can leave gaps when agents reach applications through other paths, while blanket restrictions and tool-by-tool rules make adoption harder to scale. In this discussion, Workday shares real-world lessons from rolling out AI agents to thousands of employees. Obsidian Security examines where enforcement works, where gaps remain, and how centralized policy decisions can be enforced consistently across gateways and runtimes. Attendees will learn: Why legitimate access doesn't mean every agent action should be allowed. Where gateway-only controls, blanket blocking, and tool-specific policies fall short. How centralized decisions and real-time enforcement work together across environments. How to phase AI agent governance from access and visibility to centralized control, without slowing adoption. |
|---|---|
| More info: | https://www.bankinfosecurity.com/webinars/webinar-many-paths-one-policy-enforcing-ai-agent-actions-at-runtime-w-7483?user_email=webmailbricade@gmail.com&rf=2026-11-16_ENEWS_SUB_BIS__Slot2_WEB7483&mkt_tok=MDUxLVpYSS0yMzcAAAGlhAZ3HIpqwo2tzM06EZJTy7x2e8C4Su0NzKQnxXq7umeIE3MbYUMn3ZNGb0pCm74bRhvYbRFS_vG68QdNibGWI9JsgRDliyr0OjtkrtFPeAxT-YrX0A |
| Date added | Nov. 17, 2026, 4:31 a.m. |
|---|---|
| Source | Bank Info Security |
| Subjects | |
| Venue | Dec. 9, 2026, midnight - Dec. 9, 2026, midnight |
