Enterprise AI success hinges on proprietary data and open models; only 11% prepared for agents

A new analysis of enterprise AI trends finds that proprietary data and open models are key to ROI, but only 11% of organizations are prepared for AI agent deployment. New data shows only 17% have deployed agentic AI, with 40% of projects expected to be cancelled. The governance gap and pilot-to-production failure rates remain critical barriers, reinforced by the DSG Atlanta AI Forum's focus on moving from pilots to production. Nutanix's MCP server for hybrid cloud agentic AI with RBAC and HITL directly addresses enterprise governance needs.