Join us for a live webinar on Wednesday, 8 July, 2026 at 11:00am PT (2:00pm EST 8:00pm CET).
In this interactive session Nate Demuth, Chief Architect at Manifold, will explore why blocking AI adoption is no longer a defensible position for security leaders, and what the real attack surface looks like when agents reach every desktop. Covering everything from shadow deployments to supply chain risk.
You will leave with a practical framework for governing AI agent deployment, covering policy definition, runtime observability, detection and response, and the KPIs your business will hold you to.
What the masterclass will cover
The 45-minute live session is built around the tension every security team is navigating right now: how to say yes to AI adoption and still control the risk.
The session will cover four areas. First, the business pressure: why slow security review has become indefensible and how the CISOs team are now measured on authorisation speed, not just risk prevention. Second, the real attack surface you’re actually approving when you approve an agent: architecture, permissions, supply chains, MCP gateways, and the shadow agents and personal licences that sanctioned toolchains don’t cover. Third, a defence framework that includes KPIs worth tracking, policy design from global baselines to department-level controls, runtime observability across sanctioned and unsanctioned agents, and detection and response for the new insider threat model that agents create. And fourth, what a credible authorization SLA actually looks like and what it signals to the rest of the business.
If you’re a security leader trying to figure out how to enable AI adoption without building a review bottleneck or absorbing unlimited liability, this session is built for you.
Latest articles
