AI is redefining infrastructure strategy. As AI workloads grow, many organisations are reassessing where those workloads should run, introducing private infrastructure alongside public cloud environments to help optimise AI investments. 

That focus on AI cost management is becoming increasingly important, with 98% of organizations now actively managing AI spend, up from just 31% two years ago. But the reality is more complicated. Introducing AI workloads to private infrastructure doesn't make costs disappear. It simply shifts them into different areas, from GPU investments and storage requirements to licensing, operations, and ongoing management. The true cost of AI isn't determined by where it runs, but by how well it's governed, measured, and optimized.

AI is creating a new FinOps challenge

After years of cloud optimization, many organizations have already captured the most obvious cost savings. AI is changing the rules — and with new consumption patterns and costs that traditional cloud optimization wasn't built for.

As a result, the conversation is shifting. Organizations are no longer focused solely on reducing cloud costs. They're focused on understanding, governing, and optimizing AI consumption. In fact, nearly every organization now manages AI spend, making AI cost management the top skill FinOps teams need to build and the leading priority for the future of FinOps.

The renewed interest in private cloud in the FinOps conversation.

 AI is driving many organizations to take a renewed interest in private cloud. Rising GPU demands, data residency requirements, predictable long-term workloads, and concerns about hyperscaler costs are pushing private infrastructure higher on the agenda. 

This push is reflected in FinOps adoption, with private cloud management seeing significant growth over the past year. The real challenge is understanding the full economics of AI workloads and ensuring they are deployed in a way that delivers the best value.

The hidden costs multiply 

As organizations scale AI, workloads become more spread across public cloud, private cloud, SaaS AI platforms, data platforms, and edge environments. While this provides flexibility, it also makes costs harder to track. Different platforms, teams, and billing models can create visibility gaps, making it difficult to get a clear picture of where AI spending is going.

Many of the biggest AI cost drivers are easy to overlook. Unused AI resources, orphaned storage, inefficient prompts, unused tokens, overallocated services, poor tagging, and unclear ownership can all quietly increase costs over time. As cloud and AI spending grows, even small inefficiencies compound rapidly and become significant sources of waste. Without strong visibility and governance, these hidden costs can quickly erode the value of AI investments.

Why AI demands a new approach to FinOps

GPU usage, token consumption and model operations create layers of complexity when it comes to calculating AI spending. The real challenge is understanding who's spending it and what business value it delivers.

Visibility is now the foundation of AI cost management. According to the State of FinOps 2026 report, organizations struggle most with understanding AI costs, allocating spend, and measuring ROI. Ultimately, can't optimize what you can't see.

The real question isn’t about AI spend — it’s about value.

AI is becoming part of a much larger technology investment strategy. To understand costs and outcomes, organizations need visibility across AI, cloud, SaaS, licensing, and private infrastructure. Bringing these areas together helps leaders make smarter investment decisions.

Managing technology spend is important, but measuring value is becoming the bigger priority. As AI investments grow, organizations want to know what's working, where they're seeing returns, and which initiatives deserve additional funding. The companies that can connect AI spending to business outcomes will be best positioned to maximize value.

The future of AI infrastructure is hybrid

The future of AI infrastructure is hybrid by design. Organizations are increasingly combining public and private cloud environments with SaaS and data platforms to create flexible, scalable foundations for AI-driven growth.

Competitive advantage will come from knowing where workloads belong and how to manage them efficiently. Those that can optimize costs, measure outcomes, and align infrastructure decisions with business priorities will unlock the greatest value from AI.

Take control of your hybrid cloud journey

Managing a hybrid cloud environment doesn't have to mean sacrificing visibility, control, or efficiency. With CloudHealth, organizations can confidently govern resources, optimize cloud spending, and align operations with business goals from a single platform. Reach out to our team to discover how CloudHealth can help unlock more value from your hybrid cloud investments.

Source
© 2026 FinOps Foundation Project a Series of LF Projects, LLC. State of FinOps 2026 Report