Uploaded June 2026 | Updated September 2026, 1 week ago
AI agents are taking action across the enterprise — transferring funds, refunding orders, and calling downstream tools. But when an agent triggers a backend system, who is actually authorizing the transaction?
Join Vijay Pandiarajan, VP of Product Management at MuleSoft, in this whiteboard session as he breaks down a critical security vulnerability in enterprise AI: the Identity Gap.
In this video you'll learn:
* The Identity Gap: What it is and why it stalls enterprise AI projects from moving into production.
* On-Behalf-Of (OBO) Tokens: How they protect data using a Zero-Trust, least-privilege model powered by MuleSoft Omni Gateway.
* In-Band vs. Out-of-Band Auth: The difference between In-Task Authorization and OIDC-CIBA approvals for background tasks.
* End-to-End Security: How to securely pass user identity all the way to headless Model Context Protocol (MCP) servers.
* Trusted Agent Identity: How establishing clear identity paths creates full auditability, accountability, and compliance.
If your organization is scaling an agentic workforce, security and verifiable identity are the absolute foundations of trust.
👉 Learn more about MuleSoft Omni Gateway and Trusted Agent Identity: muley.cc/43SECmw
About MuleSoft:
Connect systems, data, and AI with MuleSoft, the integration and automation foundation for the agentic enterprise. As part of Salesforce, MuleSoft enables organizations to securely connect applications, APIs, and data so AI agents can discover capabilities, take action across systems, and operate with enterprise-grade governance and visibility.
Follow us!
• Facebook: facebook.com/mulesoft
• Twitter: twitter.com/MuleSoft
• LinkedIn: linkedin.com/company/mulesoft
• Instagram: instagram.com/mulesoft
Copyright © MuleSoft
AI agents are taking action across the enterprise — transferring funds, refunding orders, and calling downstream tools. But when an agent triggers a backend system, who is actually authorizing the transaction?
Join Vijay Pandiarajan, VP of Product Management at MuleSoft, in this whiteboard session as he breaks down a critical security vulnerability in enterprise AI: the Identity Gap.
In this video you'll learn:
* The Identity Gap: What it is and why it stalls enterprise AI projects from moving into production.
* On-Behalf-Of (OBO) Tokens: How they protect data using a Zero-Trust, least-privilege model powered by MuleSoft Omni Gateway.
* In-Band vs. Out-of-Band Auth: The difference between In-Task Authorization and OIDC-CIBA approvals for background tasks.
* End-to-End Security: How to securely pass user identity all the way to headless Model Context Protocol (MCP) servers.
* Trusted Agent Identity: How establishing clear identity paths creates full auditability, accountability, and compliance.
If your organization is scaling an agentic workforce, security and verifiable identity are the absolute foundations of trust.
👉 Learn more about MuleSoft Omni Gateway and Trusted Agent Identity: muley.cc/43SECmw
About MuleSoft:
Connect systems, data, and AI with MuleSoft, the integration and automation foundation for the agentic enterprise. As part of Salesforce, MuleSoft enables organizations to securely connect applications, APIs, and data so AI agents can discover capabilities, take action across systems, and operate with enterprise-grade governance and visibility.
Follow us!
• Facebook: facebook.com/mulesoft
• Twitter: twitter.com/MuleSoft
• LinkedIn: linkedin.com/company/mulesoft
• Instagram: instagram.com/mulesoft
Copyright © MuleSoft
