- CIOsurge
- Posts
- Cloud spending surges as AI demand reshapes hyperscaler strategy
Cloud spending surges as AI demand reshapes hyperscaler strategy
AI Shifts Bottlenecks from Access to Governance, Cloud Spend Surges with AI Demand, Lower Costs Drive Higher AI Usage
Zack Tembi
March 31, 2026
AI Shifts Bottlenecks from Access to Governance, Cloud Spend Surges with AI Demand, Lower Costs Drive Higher AI Usage
Powered by Single Fin
Welcome to this week’s edition of CIOsurge!
This week:
In my conversation with Lloyd Fobi, one point stood out clearly: AI can remove reporting bottlenecks fast, but the real challenge often reappears upstream in data access, permissions, and governance.
Global cloud infrastructure spending hit $110.9 billion in Q4 2025 as hyperscalers expanded AI capacity, reinforcing how deeply enterprise AI roadmaps now depend on cloud scale.
Even as AI inference costs are projected to fall sharply, enterprise spending is still likely to rise as organizations push into more complex, token-intensive use cases like agentic AI.
Let’s make this week a game-changer.
Stay sharp. Stay ahead.
💡 Guest Expert Insights: Lloyd Fobi
AI Removes One Bottleneck and Exposes Another
AI can eliminate reporting delays, but often shifts the constraint to data access and control.
I recently spoke with Lloyd Fobi, founder of Health Scale Partners and a longtime healthcare CIO and COO advisor, and he shared a concrete example from his time as a CIO, where his team intentionally moved out of the reporting business. They structured supply chain data outside of the ERP and layered a conversational interface on top so operators could query it directly.
The result was speed. What once took weeks to answer dropped to minutes, and IT was no longer the gating function for basic operational insight.
But the more important observation is what followed. The bottleneck did not disappear. It moved. Once access expanded, the new constraint became governance. Who should have access to what data, and how should that access be controlled? AI reduced friction in one area and exposed a more sensitive control problem upstream.
That is a pattern worth paying attention to. These systems rarely simplify the environment overall. They tend to shift complexity into areas that are harder to manage if you are not prepared for them.
- Zack Tembi
Cloud spending surges as AI demand reshapes hyperscaler strategy
Global cloud infrastructure spending hit $110.9 billion in Q4 2025, up 29% year over year, as hyperscalers scaled AI capacity to meet rising enterprise demand, according to Omdia. AWS, Azure, and Google Cloud all saw strong growth, with AI driving broader infrastructure needs across compute, storage, and networking despite constraints in data center expansion.
We are watching cloud become the operational backbone of AI in real time. The scale of investment is massive, but so is the dependency. When three providers are spending hundreds of billions to keep up with demand, it tells you how critical infrastructure has become to every enterprise roadmap.
For CIOs, the question is shifting from whether to adopt cloud for AI to how to do it sustainably. Capacity constraints, vendor lock-in, and rising costs are all converging at once. The winners will be the organizations that can embed AI into workflows while maintaining flexibility across providers and keeping a tight handle on usage and governance.
- Zack Tembi
AI costs drop sharply but enterprise spend still rises
Gartner predicts AI inference costs could fall more than 90% by 2030, with models becoming up to 100x more efficient. However, enterprises are unlikely to see savings as demand shifts toward more complex, token-intensive applications like agentic AI, which drive overall spending higher despite lower unit costs.
Lower costs don’t automatically translate to lower budgets. As inference becomes cheaper, the natural response is to do more with AI, not spend less. More advanced use cases, especially agentic systems, quickly absorb those gains and push total costs back up.
This puts pressure on CIOs to stay disciplined. The real challenge is not chasing the lowest cost per token, but aligning spend with business value. Without clear prioritization, organizations risk scaling AI usage faster than they can justify its return.
- Zack Tembi
🗞️ At A Glance
💡 CIO Spotlights
RGP CIO Prashant Lamba tasked with unifying tech and scaling operations
Prashant Lamba joins RGP as CIO to lead enterprise tech strategy, focusing on a seamless digital ecosystem across consulting and on-demand services.
Lamba aims to connect systems, data, and workflows to improve speed, scalability, and client delivery.
With prior roles at Oracle and Ankura, he brings experience in turning AI and automation into practical business outcomes.
Kootenai Health elevates Todd Holling to CIO after interim role
Todd Holling, now permanent CIO at Kootenai Health, steps in after serving as interim since September 2025.
With 20+ years in healthcare IT, Holling has led major initiatives including cloud migrations, ERP rollouts, and EMR optimization.
Known for building strong IT teams, he focuses on using technology to improve clinical and operational outcomes.
Reply