Everyone is running AI. Almost nobody is governing it.
Kong, F5, Workiva and Snowflake each published AI research in the past two months. Read side by side, they describe the same gap.
TechPulse Research Desk
Report · from Kong
The State of AI Governance
How enterprise AI adoption is outpacing the controls meant to govern it, drawn from anonymised production traffic across hundreds of organisations using Kong AI Gateway.
- Format
- Report
- Published by
- Kong
- Category
- Artificial Intelligence
- Added
- 28 Sept 2026
About this report
The State of AI Governance: The Enterprise Governance Gap – Trillions of Tokens, Zero Governance explores how rapidly enterprise AI adoption is outpacing the controls needed to govern it. Based on anonymized production traffic from hundreds of organizations using Kong AI Gateway, the report provides a production-level view of how enterprises are consuming, routing, and governing AI.
Discover key insights into multi-model AI adoption, model and provider usage, industry-specific AI trends, governance gaps, and the emerging agent economy. The report highlights that 62% of organizations run two or more AI models in production, while less than 1% of organizations in the data have AI-specific governance controls in place.
Learn what organizations with AI governance in place are doing differently, how prompt-injection guards and request/response controls are being applied, and why establishing governance at the AI traffic layer is becoming increasingly important as AI agents and workloads scale.
Listen or skim
Summaries are generated from this page’s text. Read the full piece for the detail.
Our take
The newsroom read this
More from Kong
All 2 →
Elsewhere in Artificial Intelligence
See the category →Toolkit
AI Governance Toolkit
Building AI governance one policy at a time invites gaps—and by the time you find them, the damage is done. A more effective path is to start deliberately, prove value quickly, and scale from there.
WorkivaAI
Toolkit
Is Your Audit Testing Ready for AI Automation?
Where can AI create meaningful value—and where does it introduce more risk than benefit? The answer depends on the activity.
WorkivaAI
Toolkit
How ready is your organization for AI? But really?
The pressure to adopt AI is immense, but rushing deployment without measuring your organization’s true readiness is a recipe for governance failure and unchecked risk.
WorkivaAI