For years, the growth of the public cloud rested primarily on migrating enterprise applications and infrastructure.
Recent quarterly results from Amazon, Microsoft, and Alphabet show that the engine has shifted. Generative AI has become the main driver of cloud resource consumption, whether for training models, running inferences, or deploying AI agents.
But while all three hyperscalers are accelerating their cloud activity, their trajectories diverge. Microsoft maintains the highest growth, AWS regains a pace it hasn’t seen in years, and Google Cloud continues its ascent, backed by the adoption of Gemini and Vertex AI.
AWS regains nearly 40% growth
Amazon Web Services posts a quarterly revenue of $42.2 billion, up 37% year over year. It marks AWS’s strongest growth in years.
For Andy Jassy, Amazon’s chief executive, the acceleration is directly tied to enterprise investments in generative AI. The group highlights momentum in the adoption of Amazon Bedrock, demand for its Trainium chips, and the increasing use of GPU infrastructures dedicated to inference.
The CEO notes that demand still exceeds current capacity and confirms that part of the compute capacity planned for 2027 is already reserved by customers.
To support this demand, Amazon now plans $220 billion in capital expenditures in 2026, largely aimed at building new data centers and AI-focused compute capacity.
Azure keeps the strongest momentum
Microsoft sustains the strongest cloud growth among the major hyperscalers.
Azure posts a YoY growth of 43% while Microsoft Cloud overall reaches $59.3 billion in quarterly revenue, up 27%.
The group attributes this performance to the broader diffusion of its AI offerings, notably Azure AI and Microsoft 365 Copilot. Microsoft now notes more than 30 million paying Microsoft 365 Copilot users.
Another highlighted indicator: Azure now generates more than $100 billion in annual revenue, underscoring the platform’s scale-up.
Despite the magnitude of investments in AI infrastructure, Microsoft keeps its $175 billion investment program for fiscal 2026.
Google Cloud accelerates again
Google Cloud likewise confirms a clear acceleration in its growth. For its second fiscal quarter, the division generated $24.8 billion in revenue, up 82% year over year. Its operating income reached $8.8 billion, up from $2.8 billion a year earlier, while its order backlog grows to $514 billion.
The group highlights the adoption of Gemini Enterprise, now used by almost 90% of Fortune 100 companies, along with the progress of Vertex AI and Google Cloud Platform (GCP) AI-specific infrastructures.
Alphabet has raised its 2026 outlook, now planning capital expenditures of between $195 billion and $205 billion primarily to develop its data centers and AI compute infrastructures.
This quarter confirms Google Cloud’s scale-up. Long outpaced by AWS and Azure, the third-largest hyperscaler now displays the fastest growth in the market.
A competition shifting toward execution capacity
The three providers argue that compute capacity is becoming as differentiating a factor as AI models themselves. Their results illustrate this evolution: cloud is becoming the enterprise AI execution platform, combining compute power, language models, development tools, inference services, and applications.
For CIOs, this shift could have direct consequences for AI project roadmaps. Available capacities, resource provisioning timelines, and contractual commitments with cloud providers may become as important selection criteria as platform features.
