OPAQUE AI Confidential Webinar: From Trust to Proof - The Next Generation of AI Agent Governance

AI agents are rapidly evolving from assistants into autonomous systems capable of accessing sensitive, proprietary enterprise data, executing workflows, approving transactions, and coordinating with other agents. As organizations adopt these systems, a new question emerges: How do you prove an AI agent actually operated according to policy, not simply trust that it did? In this webinar, Imran Siddique, Chief Platform Officer at OPAQUE and creator of the open-source Agent Governance Toolkit (AGT), joins Jason Lazarski, Director of Growth at OPAQUE, to discuss why AI governance must evolve from policies and audit logs to cryptographic proof backed by confidential computing and open standards. Learn how organizations can control AI agents, verify what they actually did, and build trustworthy AI systems that are ready for enterprise and regulated environments. YOU'LL LEARN: • Why governing AI agents is fundamentally different from governing traditional software • The difference between controlling an AI agent and proving you controlled it • Why open standards like Agent Manifest, Confidential MCP (cMCP), and TRACE are becoming the foundation for trustworthy AI • How confidential computing and hardware attestation provide verifiable evidence—not just logs • Where to start today with AGT and Confidential MCP • Why governance doesn't have to sacrifice performance THE OPEN STANDARDS DISCUSSED IN THIS SESSION: • Agent Manifest — binds the 10 artifacts that define an agent to hardware at deployment https://github.com/agentrust-io/agent-manifest • Confidential MCP (cMCP) — MCP gateway running inside a TEE, with Cedar policy measured into hardware attestation before any code runs https://github.com/agentrust-io/cmcp • TRACE — open attestation standard; a signed trust record per agent run, verifiable offline without calling the operator https://github.com/agentrust-io/trace-spec • Agent Governance Toolkit (AGT) — the open-source runtime governance layer https://github.com/microsoft/agent-governance-toolkit • Runnable examples across all three: https://github.com/agentrust-io/examples • AgenTrust Fellowship — applications open August 1 to August 29, 2026 https://agentrust.io • Agent Control quickstart is live at quickstart.opaque.co Whether you're building AI agents, securing enterprise AI, or preparing for production deployments, this session offers a practical roadmap for moving from trust to proof. If you’re building, deploying, or governing AI systems, you won't want to miss this. OPAQUE is the Confidential AI company. To learn more about OPAQUE's Confidential AI Platform, visit https://www.opaque.co.