Cloud Infrastructure Spend Hits $110B: How Hyperscalers Are Building for AI at Scale

The final quarter of 2025 marked a turning point for the global cloud industry. According to a research report by Omdia, global spending on cloud infrastructure reached $110.9 billion, reflecting a 29% year-over-year increase.

For the world’s leading hyperscalers, AWS, Microsoft, and Google, this growth signals a shift as enterprise AI moves from experimentation into production environments. The report found that these three providers accounted for 66% of total cloud infrastructure spending in Q4 2025, underscoring the concentration of market power at the top.

AI Infrastructure Spending Surges as Hyperscalers Scale Capacity

To meet rising demand, hyperscalers are stepping up capital investment significantly. Rather than short-term expansion cycles, much of this buildout appears linked to ongoing AI-driven demand that is expected to persist.

• AWS expects its capital expenditure to reach approximately $200 billion in 2026, more than 50% above 2025 levels.
• Microsoft reported quarterly capital expenditure of $37.5 billion, up by nearly $15 billion year over year. Google, meanwhile, raised its 2026 capital expenditure guidance to between $175 billion and $185 billion, more than double the prior year’s level.

Rachel Brindley, senior director at Omdia, said, “For cloud vendors, the challenge is no longer just about scaling capacity quickly enough to meet surging demand, but about doing so with discipline across investment pace, resource allocation, and global operational efficiency. As AI continues to raise infrastructure requirements while constraints remain, vendors that can expand in a more targeted and efficient way will be best positioned to lead in the next phase of competition.”

From a hyperscaler standpoint, this level of spending reflects changing demand patterns. AI workloads are not limited to specialized compute, they are also increasing requirements across storage, networking, and general-purpose infrastructure.

Hyperscalers Shift Focus to Operationalizing AI in Enterprise Workflows

Competition is beginning to extend beyond infrastructure toward how AI capabilities are applied in enterprise environments. Increasingly, attention is turning to embedding AI agents into business processes and operational workflows, although adoption maturity will vary by organization.

“For enterprise customers, the key question is whether these capabilities can be embedded into existing systems, workflows, and data environments, and then scaled reliably in production,” said Yi Zhang, senior analyst at Omdia. “This is pushing cloud vendors to invest more heavily in tool governance, workflow orchestration, and deployment capabilities, helping AI move closer to operational use at scale.”

In practice, hyperscalers are moving beyond model access to delivering integrated capabilities that support enterprise use cases:

• AWS introduced agent-focused offerings such as Kiro and Amazon Q. Its Nova platform enables enterprises to incorporate proprietary data into foundation models.
• Microsoft Azure is advancing “agentic” operations, extending Copilot into optimization and application modernization workflows.
• Google Cloud is focusing on agent governance through Vertex AI and expanding partnerships to support model development and deployment.

Cloud Market Growth Highlights Backlog and Demand Signals

The Q4 2025 results also point to differences in how hyperscalers are converting demand into revenue, with backlog emerging as a key indicator of future activity:

• AWS remains the market leader with a 32% share, reporting a total backlog of $244 billion.
• Microsoft Azure captured a 22% share with 39% year-over-year growth.
• Google Cloud delivered a 50% growth rate, bringing its market share to 12% and increasing its backlog to $240 billion.

For enterprise buyers, backlog reflects committed demand that has yet to be fulfilled. This can indicate pipeline strength, longer-term contracts, and continued demand for infrastructure capacity, although delivery timelines and constraints may vary.

Efficiency and Execution Define the Next Phase of Cloud Growth

As the industry looks toward a forecasted 27% growth in 2026, according to Omdia, the challenge for hyperscalers is evolving. It is no longer only about scaling capacity quickly; attention is also shifting to how efficiently that capacity is deployed and managed.

In an environment where constraints remain, vendors that take a more targeted and disciplined approach to infrastructure expansion may be better positioned to support enterprise AI workloads over time.

Recently, Microsoft said it is validating an NVIDIA-powered AI supercomputer in its Azure environment, one of the first hyperscaler deployments of this scale. The move signals a move from incremental capacity adds to purpose-built, large-scale systems designed specifically for sustained AI workloads, reinforcing how infrastructure design itself is becoming a differentiator.