# How to Achieve AI and API Governance at Enterprise Scale

This talk is about how APIs are becoming the foundation for governing, securing, and scaling the fast-rising world of AI agents, models, and machine-to-machine traffic. The speaker argues that the agentic era is arriving faster than expected, citing examples such as two new personal AI agents reaching 3 million downloads in 12 days, one already driving a billion-dollar e-commerce run rate, and Kong AI gateway traffic growing from about 100 billion tokens a year ago to 32 trillion, a 30x increase. He warns that this shift brings new risks around security, runaway token spending, shadow AI, regulation, and agent payments, with HTTP 402 re-emerging for paywalled API access. At the same time, enterprises are reducing inference costs through model routing, semantic caching, and growing use of open-weight models. The core message is that API platform teams already know how to manage this transition and should unify API, event, model, MCP, and agent traffic through a single governance and control layer.