Cloud computing reshaped how organizations build, deploy, and scale digital services, and it’s been the backbone of transformation for nearly two decades. But as 2025 draws to a close, something new is happening. Cloud computing has reached maturity; I’d argue it’s even reached saturation.
Every enterprise operates in some hybrid or multi-cloud form. Costs are rising while SaaS fatigue is setting in. AI is rewriting infrastructure economics and sustainability concerns are mounting.
The cloud isn’t fading away, though. It’s evolving into something broader: a distributed fabric of compute that stretches from hyperscale data centers to on-premises clusters and the edge of networks and devices. The next era of IT will not be defined by where workloads run, but by how intelligently they move.
Peak cloud and the hybrid reality
A decade ago, cloud migration was a mandate. We all saw numerous boards ask, “How fast can we move?” Today, though, that question has changed to, “What should stay and what should come back?”
Enterprises have learned that not all workloads belong in public clouds. Performance, cost and compliance simply vary too widely. The result is a deliberate hybrid reality where compute is distributed by design.
From what I’ve observed, nearly all large enterprises now operate across three or more cloud providers, emphasizing interoperability and cost control as top priorities. Cloud has become the default fabric of IT, but that means it’s no longer a differentiator.
I would assert that the strategic shift here is from migration to optimization. CIOs’ focus now lies in orchestrating across platforms, negotiating value and deciding which workloads create the most impact in each environment.
To put it more succinctly, I would argue that the cloud conversation has matured from “How do we get there?” to “How do we get smarter about what runs where?”
AI and the new compute gravity
Artificial intelligence has introduced a certain gravity to cloud computing. AI workloads are massive, power-hungry, and location-sensitive. They pull data and compute closer together and, as I’ve seen with many of my clients, reshape entire data centers’ economics.
Meanwhile, cloud providers are no longer just service vendors; they’re infrastructure engineers designing GPUs, AI-specific chips and advanced cooling systems. Enterprises are beginning to mirror that behavior at a smaller scale, building private GPU clusters to control costs and manage sensitive data.
The boundary between cloud, hardware and AI is fast becoming a mirage. Compute has become a strategic resource instead of a commodity. For CIOs, this means thinking beyond “cloud services” to a new model of compute availability: the ability to run intelligent workloads wherever they deliver the most value.
In this environment, power, proximity and compute performance are the new currencies of cloud strategy.
The SaaS slowdown and rise of consumption models
As we’ve watched the cloud mature, SaaS has become both indispensable and utterly exhausting. Subscription models once promised predictability, but they’ve now created renewal fatigue. Enterprises pay again and again for capabilities they rarely use, while per-user licensing and seat expansion inflate costs each year.
At the same time, AI services are introducing a different pricing paradigm: pure consumption. Instead of fixed subscriptions, usage is measured per inference, per token and per API call. Organizations pay only for what they consume and see direct correlation between cost and compute value.
I believe that this will be the dominant model for software in the decade ahead. It’s a paradigm that aligns incentives for both customers and providers while rewarding efficiency and transparency. However, it also demands new financial governance from IT leaders.
In the coming years, CIOs will need real-time visibility into variable spend, predictive budgeting tools and procurement models built around outcomes instead of licenses. Software will no longer be something you own; it will be something you continually earn through value delivered.
Edge computing and the rise of the micro-cloud
At the same time, compute is moving to the edge. As IoT, analytics and AI converge, organizations can no longer afford the latency or bandwidth of sending every transaction back to a “hyperscaler” region.
Edge computing brings processing closer to where data originates: on factory floors, in hospitals, at retail stores or even within telecom towers. Verizon, AT&T and other providers are investing heavily in distributed “micro-clouds” that operate at network edges, which enable real-time decisioning for connected systems.
Enterprises are following suit, deploying localized compute infrastructure inside their own facilities to handle high-volume or sensitive workloads. These micro-clouds act as intelligent satellites in that they’re connected to the larger ecosystem, but optimized for speed, sovereignty and control.
For CIOs, this distributed architecture introduces new considerations: how to secure and manage assets across hundreds of locations, how to unify governance and how to ensure resilience when workloads shift between core and edge. The infrastructure of the future will look less like a data center and more like a constellation of compute nodes.
Sustainability and the power problem
AI’s explosive growth has exposed compute’s physical limits. Data centers that once measured efficiency in kilowatts are now measuring in megawatts. Water usage and carbon footprints have become board-level topics.
The environmental cost of intelligence is real. A single large-scale AI training run can consume as much energy as hundreds of homes use in a year. For technology leaders, this changes the definition of scalability.
The next wave of innovation will hinge on efficiency: new cooling methods, renewable microgrids and smarter workload distribution that minimizes compute waste. In many organizations, sustainability officers are now sitting alongside CIOs to shape infrastructure decisions.
This tells me that the future of cloud strategy will be measured not ‘only’ in uptime or cost, but in carbon per compute. The enterprise that can do more with less energy will have both an operational and reputational advantage.
Leadership in the post-cloud era
The CIO’s role is evolving even faster than usual!
In the first era of cloud, leadership meant driving migration. In the second, it meant managing cost and security. In this third era, where compute reigns supreme, leadership means optimization and orchestration.
The modern CIO must balance five dimensions:
- Economics: consumption vs. commitment
- Performance: proximity vs. scale
- Security: control vs. agility
- Sustainability: efficiency vs. expansion
- Innovation: stability vs. experimentation
This is fundamental strategic governance, not some siloed balancing act. It requires partnership with finance, operations, sustainability and product teams to ensure that compute decisions align with business outcomes. It reflects the reality that transformational success is one motion and can’t happen in a fragmented paradigm.
As AI accelerates, this orchestration mindset will become essential. Not every model needs to live in the cloud, but every enterprise must learn how to manage a distributed fabric of compute that spans it.
The cloud is everywhere, and so is opportunity
Cloud computing is not disappearing; it’s decentralizing. The boundaries between public, private and edge are dissolving into a single continuum of compute.
For CIOs, this evolution brings both complexity and opportunity. The challenge is no longer simply to migrate, but to design architectures that can adapt: placing intelligence, data and compute exactly where they create the most value.
Ultimately, the future of cloud is about liberation: the ability to run, learn and scale anywhere. The enterprises that master this flexibility will define the next era of digital leadership.
This article is published as part of the Foundry Expert Contributor Network.
Want to join?
Read More from This Article: After the cloud: The future of compute is everywhere
Source: News