How Enterprises Lose Control as AI Scales

AI adoption is accelerating faster than most organisations can control and the result is growing chaos. In this episode of IoT and AI Leaders, Nick Earle is joined by Santosh Kaveti, CEO of ProArch, and Jim Spignardo, Director of Cloud Strategy and AI Enablement, to explore why ‘AI-first’ thinking without foundational discipline is creating serious operational, security, and governance risks. Rather than chasing hype, ProArch argues for a foundation-first approach: slow down, define processes, establish governance, and build trust before embedding AI into workflows, especially as IoT expands the data surface dramatically. The conversation covers: Why cheap, accessible AI tools are amplifying organisational chaos How weak processes and governance cause most AI failures Why regulated industries are feeling the pressure first How IoT could increase enterprise data sources by 50x Why boards must take accountability for AI risk and reputation What ‘operationalising AI’ means beyond chatbots Why trust, auditability, and explainability matter more than intelligence Episode Highlights 00:02:15: Introduction to ProArch 00:04:30: Realising the potential of AI 00:06:45: What are the 'unknown unknowns' of AI? 00:12:20: The "Donald Rumsfeld era" of AI 00:18:00: ERP/MRP analogy 00:22:15: Balancing visibility and control with innovation 00:28:45: The "trust layer" or "truth layer" 00:35:30: Real-world power plant example 00:42:15: Ontology mapping 00:48:30: Three board-level risks identified 00:52:00: Final takeaways