According to data from Omdia, global spending on cloud infrastructure services reached $110.9 billion in the fourth quarter of 2025, representing a 29% increase compared to the same period last year. The growth rate accelerated further from the previous quarter, marking the sixth consecutive quarter of market expansion exceeding 20%. As enterprise demand for AI shifts from experimental phases to production deployment, leading cloud providers (hyperscalers) are continuing to increase investments to expand AI infrastructure capacity. Looking ahead to 2026, Omdia forecasts that global cloud infrastructure service spending will grow by 27%.
Future market differentiation will increasingly depend on infrastructure scale, capital efficiency, and the maturity of AI agent-related platform capabilities. In the fourth quarter of 2025, Amazon Web Services (AWS) saw its growth accelerate to 24%, while Microsoft Azure and Google Cloud achieved year-on-year growth of 39% and 50%, respectively. AI demand is no longer limited to specialized computing power like GPUs but is also driving growth in broader infrastructure needs, including CPUs, storage, and networking.
Enterprises are accelerating the development of AI applications around agents, workflows, and data integration, leading to sustained growth in demand for integrated cloud environments with orchestration, scalability, and governance capabilities. This further solidifies the core role of cloud platforms as the foundation for AI and is driving the migration of more traditional and emerging workloads to the cloud.
Concurrently, AWS, Microsoft Azure, and Google Cloud all reported growing order backlogs, reflecting continued strong demand and increasing corporate willingness to invest in AI and cloud infrastructure. Rising demand is also prompting leading cloud providers to further increase capital expenditures to support AI infrastructure expansion. In 2026, AWS is projected to reach capital expenditures of $200 billion, a more than 50% increase from approximately $132 billion in 2025. Microsoft reported quarterly capital expenditures of $37.5 billion, an increase of nearly $15 billion year-on-year. Google raised its 2026 capital expenditure guidance to between $175 billion and $185 billion, representing a doubling from the previous year.
Rachel Brindley, Senior Director at Omdia, stated: "For cloud providers, the challenge is no longer just about rapidly scaling capacity to meet surging demand, but also about achieving a more disciplined balance between investment pace, resource allocation, and global operational efficiency. AI continues to drive infrastructure demand higher, yet various constraints remain unresolved. In this context, providers that can expand more efficiently and with greater focus will secure a more advantageous position in the next phase of competition."
Meanwhile, competition is extending beyond model access and infrastructure scale to the application layer, particularly in the development and deployment capabilities of AI agents. Omdia Senior Analyst Yi Zhang commented: "For enterprise customers, the decisive factor is whether these capabilities can integrate into existing systems, workflows, and data environments, and operate stably at scale in production settings. This trend is driving cloud providers to increase investments in tool governance, workflow orchestration, and deployment capabilities to accelerate the scalable implementation of AI."
For example, AWS has launched several productized intelligent agent solutions, such as Kiro, Amazon Quick, Transform, and Connect, while Microsoft is extending agent capabilities to scenarios like cloud operations and application modernization.
AWS maintained its leading position in the global cloud infrastructure market in Q4 2025, with a market share of 32%. Its revenue grew 24% year-on-year, an improvement from the previous quarter. By the end of the quarter, AWS's order backlog reached $244 billion, highlighting sustained strong market demand. AWS indicated that Amazon Bedrock has achieved an annualized revenue run rate in the billions of dollars, with customer spending increasing 60% quarter-on-quarter.
In December 2025, AWS launched Nova Forge, enabling enterprises to incorporate proprietary data during the early training stages of Amazon Nova models to build customized foundation models (Novellas), thereby supporting deep model customization for enterprise-grade AI agents. AWS also introduced a series of productized agent solutions, including Kiro, Amazon Quick, Transform, and Connect, to accelerate the translation of model capabilities into tangible business value. At the infrastructure level, AWS continues to expand its global footprint, increasing data center capacity investments in Europe and the United States to meet growing AI computing demands.
Microsoft Azure retained its position as the world's second-largest cloud service provider in Q4 2025, with a 22% market share and revenue growth of 39% year-on-year. Since December 2025, Microsoft has continued to expand the model ecosystem of Azure AI Foundry, adding models such as Mistral Large 3, GPT-5.2, and Claude Opus 4.6, further strengthening its position as an enterprise multi-model AI platform. Concurrently, Azure is extending agentic AI from model access to the enterprise execution layer. Agentic cloud operations capabilities launched in February 2026 extended Azure Copilot to cloud operations and continuous optimization scenarios, while new agent capabilities introduced in Azure Copilot and GitHub Copilot in March further integrated application modernization into end-to-end workflow systems.
On the infrastructure front, Microsoft announced in February 2026 that its Eastern Saudi Arabia data center region will officially launch in the fourth quarter of 2026, further expanding its localized cloud and AI infrastructure footprint.
Google Cloud maintained its position as the third-largest global cloud service provider in Q4 2025, achieving robust year-on-year growth of 50% and increasing its market share to 12%. By the end of the quarter, its order backlog reached $240 billion, a significant increase from $157.7 billion in the third quarter, indicating substantially improved demand visibility.
In January 2026, Google entered a multi-year collaboration with Apple to co-develop next-generation Apple Foundation Models based on Gemini models and Google Cloud technology. Since December 2025, Google Cloud has continued to enhance its enterprise AI platform Vertex AI, adding capabilities including Gemini Embedding, Gemini 3.1 Pro, and Nano Banana Pro/2, further strengthening enterprise application capabilities in retrieval, complex reasoning, and multimodal generation.
Simultaneously, Google Cloud is continuously improving the deployment capabilities of enterprise AI agents. Through tool governance features in Vertex AI Agent Builder and Provisioned Throughput's stable support for high-concurrency scenarios, it is promoting the scalable deployment of AI agents in production environments.
Comments