AI is clearly accelerating demand for cloud computing, but not in the way many expected. The biggest story today isn't software innovation alone; it's the extraordinary capital flowing into physical infrastructure to support AI at scale. Chips, networking gear, power systems, and massive data centers are becoming the strategic center of gravity for the cloud market as providers race to support model training and inference workloads.
The numbers are hard to ignore. US technology companies, including Alphabet, Amazon, Meta, and Microsoft, are expected to spend about $650 billion on AI-related infrastructure in 2026, up from roughly $410 billion in 2025, according to analysis cited by Reuters. That kind of growth tells us something important: AI is not just another software wave that sits neatly atop the existing cloud stack. It is forcing a redesign of the stack itself.
The infrastructure race behind AI
That redesign reaches deep into networking and data movement. Nvidia recently announced plans to invest $2 billion each in photonics companies Lumentum and Coherent, underscoring where the pressure points are emerging. The issue is no longer only raw compute. It is also how quickly data can move between processors, racks, and clusters without creating unacceptable bottlenecks or power inefficiencies. As AI systems scale, latency, throughput, and energy usage become first-order economic concerns.
Hyperscalers have long dominated enterprise IT by offering seemingly infinite compute, storage, and networking resources on demand. But AI workloads are different. Training large language models requires massive parallel processing, while inference demands low-latency responses and high availability. These requirements stress traditional data center architectures and force cloud providers to invest billions in specialized accelerators, high-bandwidth interconnects, and advanced cooling systems. The capital intensity of this infrastructure is reshaping the competitive landscape, benefiting those with deep pockets and scale.
Most AI starts in the public cloud
When companies are experimenting, speed matters more than optimization. Public clouds give teams immediate access to GPUs, foundation model APIs, vector databases, orchestration tools, security controls, and integration services. They also allow businesses to quickly start pilots without waiting for procurement cycles, data center expansions, or specialized infrastructure teams.
Given the high level of uncertainty, the public cloud is often the right choice for first-generation AI. Enterprises do not yet know which use cases will deliver value, how much inference traffic they will see, or which architecture model will ultimately survive. At this stage, the ability to quickly try many things is more important than squeezing every dollar from the underlying infrastructure. Managed services reduce friction, and friction is the enemy of early adoption.
This is why we are seeing strong initial demand for AI land in public cloud environments. Enterprises are building chatbots, copilots, knowledge assistants, document automation systems, and code generation tools there because the cloud dramatically lowers the barrier to entry. It provides compute as well as a full operating environment for AI experimentation, including pre-built models, data pipelines, and observability tools that accelerate development cycles.
Next-gen AI systems present choices
The second generation of enterprise AI systems looks different. Once a use case proves its value and usage becomes persistent, the financial model changes. A workload that looked inexpensive during a proof of concept can become shockingly expensive when it runs at production scale, especially if it depends on premium GPU instances, high-performance storage, constant network traffic, and managed services layered on top of one another.
That is where repatriation enters the conversation. We are starting to see a pattern in which enterprises build first-generation AI systems on public clouds, learn what works, and then move some of those workloads back on-premises or onto so-called neocloud providers that offer AI-optimized infrastructure at a lower cost. This shift is not a rejection of cloud computing but rather a maturation of AI economics.
On-premises deployment is attractive when utilization is steady, data gravity is high, governance requirements are strict, and the organization has sufficient scale to justify owning or directly controlling the infrastructure. Many large enterprises already run private data centers for mission-critical applications, and extending those environments to support AI inference can be more cost-effective than paying hyperscaler premiums for every GPU cycle. Moreover, data residency and regulatory compliance often dictate that sensitive data cannot leave corporate boundaries, making on-prem AI a necessity for some industries.
Neocloud options become attractive when enterprises still want an external provider but do not want to pay the full premium often associated with large hyperscalers. These specialized providers are increasingly positioning themselves around dense GPU capacity, simpler pricing, and architecture built specifically for AI rather than for general-purpose enterprise IT. They offer a middle path: the flexibility of the cloud with the cost discipline of dedicated infrastructure. Some neoclouds even allow bare-metal access to GPUs, which can reduce overhead and improve performance for tightly coupled training jobs.
This is an important adoption pattern because it dispels the old assumption that cloud migration is always one-way. In the AI era, workload placement is becoming more fluid. Enterprises are learning that the best place for experimentation may not be the best place for steady-state production and that AI economics can punish architectural laziness much faster than traditional enterprise applications ever did.
AI and public cloud demand
How much demand will AI drive for public cloud computing? Quite a lot, especially in the near term. Every major enterprise AI initiative will likely engage the public cloud in a meaningful way, whether for model development, training bursts, integration services, security tools, or global deployment. But it would be a mistake to assume that all demand will remain locked in traditional hyperscalers over time.
Some AI workloads will stay in the public cloud permanently because they are bursty, globally distributed, hard to predict, or tightly coupled to cloud-native services. For example, customer-facing inference endpoints that experience spiky traffic patterns benefit from the elastic scaling that hyperscalers provide. Likewise, teams that rely heavily on managed services like Amazon SageMaker, Google Vertex AI, or Azure Machine Learning may find it difficult to decouple their workflows from the broader ecosystem, making continued cloud usage the most practical option.
Other workloads, especially those with stable usage patterns and heavy inference volume, will be candidates for relocation. A vision model that processes millions of images per day, or a natural language processing pipeline that handles a predictable number of requests, can often run more economically on dedicated hardware placed either on-premises or at a neocloud provider. The key is to understand the total cost of ownership, including power, cooling, networking, and engineering time.
Economics will drive those decisions more than ideology. As CFOs scrutinize AI budgets, they will demand clearer returns on investment. The hyperscalers are not standing still; they are introducing cost-saving options such as spot instances for non-critical training runs, custom silicon like Google's TPUs and Amazon's Trainium, and multi-year commitments that lower unit prices. However, these discounts often come with lock-in, which can reduce future flexibility.
The likely outcome is a more segmented market. Public clouds will dominate the front end of AI adoption and continue to play a major role in hybrid operations. On-premises environments will regain relevance for cost-sensitive, steady-state, and compliance-heavy workloads. Neocloud providers will grow as a middle option for enterprises seeking external AI capacity without paying full hyperscaler prices. In short, AI will increase public cloud demand, but it will also heighten scrutiny of the correct fit in the long term.
Three factors to consider
First: Speed and cost are distinct metrics. The public cloud is usually the fastest way to get an AI initiative off the ground, and that speed has real business value. But the architecture that wins a pilot may end up destroying the production budget. Enterprises need a placement strategy from day one, even if they start in the cloud. This means tracking not just initial spend but also projected costs at various scale levels, and being willing to reassess as usage patterns solidify.
Second: AI workload economics differ from those of traditional applications. Training, inference, data movement, storage, and model serving can interact in ways that quickly create cost surprises. Organizations should model not only compute usage but also utilization patterns, network flows, and the costs of managed services surrounding the core AI stack. Without that discipline, they risk designing systems that are technically elegant but financially unsustainable. For instance, a model that is retrained monthly may incur significant egress fees when moving training data to the cloud, and those fees can dwarf the compute savings.
Third: Future flexibility matters more than short-term convenience. Enterprises should avoid building AI systems so tightly around a single provider's proprietary stack that moving becomes painful or impossible. The winners in this market will be the companies that preserve optionality, enabling them to shift workloads across public clouds, on-premises environments, and emerging neocloud platforms as economics, regulations, and business requirements evolve. This includes using portable formats like ONNX for models, containerization with Kubernetes, and multi-cloud data layers that abstract away provider-specific APIs.
The real question is not whether the cloud will benefit, but how long each AI workload will remain in the cloud. AI will unquestionably generate significant new demand for public cloud computing. For most enterprises, AI workloads will stay in the cloud long enough to enable rapid innovation, but they will not necessarily remain there forever. The hyperscalers have a clear advantage in the early stages of the AI lifecycle, but as the technology matures and cost pressures intensify, the ownership of AI workloads will become a more contested and nuanced landscape. The ability to move between environments will be a key competitive differentiator, and those who plan for that flexibility today will be best positioned to thrive in the AI-driven future.
Source: InfoWorld News