video1.0<iframe src="https://www.loom.com/embed/1464c2d086104a4290a7d2abe4cf160b" frameborder="0" width="1920" height="1440" webkitallowfullscreen mozallowfullscreen allowfullscreen></iframe>14401920Loomhttps://www.loom.com14401920https://cdn.loom.com/sessions/thumbnails/1464c2d086104a4290a7d2abe4cf160b-61c81820494be70f.gif241.2812In this video, I present the Policy Governed Agent Payments system designed for LabLab.ai Agentic Commerce. As AI agents autonomously handle payments, we face risks like overspending and bypassing safeguards, which is why our approach emphasizes strict PJAX principles. The AI proposes structured payment intents, but the treasury contract enforces limits and protections, ensuring that even if the AI is untrusted, losses are contained by policy. I demonstrate five scenarios to showcase our defense mechanisms, including valid payments and protections against invalid requests. I encourage you to consider how these principles can be applied in your own projects.