# Agents are arriving in every function. The question nobody wants to own is who keeps control of them.
Agents are landing in every function, but nobody owns the question of how they are governed. Watch a practical framework for who should own AI governance, where control is actually possible, and how to keep agents on a short leash without slowing builders down. Includes a live demo.
You’ll leave with:
- - A mental model of how platform, data, AI and security responsibilities fit together, and who should own what
- - A way to test any vendor’s governance claim
- - A clear approach for giving every agent an identity, scoping its access and answering “what did that agent just do?”
- - A shared definition and a short checklist of questions to start your next internal conversation from
**REGISTER NOW!**
**Title: **AI Connectivity for the Agent Era: who owns governance, and how you stay in control
**Synopsis: **Picture your company in a year. Agents everywhere, each one calling models, tools and APIs. It looks like progress until someone asks who is in charge of all of it. Ask who owns AI governance in your organisation and you will hear three answers. Platform says it is theirs. Data says it is theirs. Security says it is theirs. In practice none of them owns it, and nobody sits in the middle setting the guardrails. Meanwhile every vendor claims governance, which makes the noise worse. Underneath it sits a worry that lands at CIO level: data leaking into public models, with personal consequences for the person who signed it off.
This session gives you a way through. You have solved this problem once already. When microservices multiplied API calls, gateways put authentication, limits and logging in one place, and teams shipped faster because policy was no longer their job. Agents multiply tokens, tool calls and events. The traffic is new, but the needs are the same: identity, policy, visibility and cost. We will show where control is actually possible, how to govern agents and MCP access without slowing the teams building them, and what it looks like in practice with a short live demo of a governed agent running through Kong AI Gateway.
**Who should watch: **CIOs, heads of platform, security leaders and architects who are rolling out agents and need to know who is accountable for them.