The enterprise cloud where AI Comes Together
By Javier Surasky
The
original article was written in Spanish. This English version is a translation
prepared with ChatGPT and reviewed by the author.
.png)
Microsoft Azure enters the AI power map as an enterprise cloud platform that offers computing, storage, and networks and, through them, organizes the environment in which organizations integrate their models, data, applications, security, identity, regulatory compliance, and automation tools.
Its place
in the ecosystem is explained by the technical capacity of its data centers:
Azure is part of a broader architecture within the “Microsoft ecosystem,” which
includes Microsoft Foundry — formerly Azure AI Foundry — a
Microsoft platform for building, deploying, evaluating, governing, and
scaling AI applications and agents within Azure, as well as products such
as Microsoft 365, Copilot, GitHub, and Dynamics, among others.
This
unified enterprise ecosystem allows Microsoft Azure to operate as a
cross-cutting layer of AI embedded in documents, meetings, code repositories,
business processes, access controls, and security systems.
More
importantly, it is precisely this integrated character that sustains Azure’s
power: without needing to produce models itself, it controls the environment in
which they are deployed, connected to internal data, audited, monitored, and so
on. This places Microsoft in a space of intermediation between material
infrastructure, foundation models, and the everyday operations of companies and
public administrations.
The
financial data show the scale of that position: in its Annual Report 2025,
Microsoft reported that Microsoft Cloud revenues reached 168.9 billion dollars
in fiscal year 2025 and stated that Azure and other cloud services drove much
of the growth in server products and cloud services. The company also
acknowledged pressure on Microsoft Cloud’s gross margin due to the cost of
scaling AI infrastructure (Microsoft, 2025a).
To
understand Azure’s place, it is worth remembering that using generative AI in
an organization is not simply a matter of accessing a chatbot or subscribing to
an API. It requires deciding where data will be located, under what
permissions, which models may be invoked, what records remain available, what
security policies apply, which provider responds in the event of incidents, and
which jurisdiction conditions the processing of information.
For that
reason, Microsoft’s cloud brings together models sold directly by Azure, with
greater integration, support, and enterprise conditions, and partner or
community models, whose support and validation depend to a greater extent on
external providers that pay Microsoft for the use of space in Azure (Microsoft
Learn, 2026).
The
alliance between Microsoft and OpenAI reinforces Azure’s power: in April 2026,
Microsoft announced a modification to its agreement with OpenAI under which
Microsoft remains OpenAI’s main cloud partner and OpenAI products are launched
first on Azure, unless Microsoft cannot or decides not to support the necessary
capabilities; in return, OpenAI may serve its products to customers on any
cloud provider, and Microsoft’s license to OpenAI’s intellectual property
becomes non-exclusive until 2032 (Microsoft, 2026a).
From a
power perspective, this reveals a double dynamic: Azure remains a privileged
infrastructure for OpenAI’s AI, while Microsoft reduces the risk of depending
on a single partner by expanding its model catalogue. In other words, Azure
strengthens its position as a provider of computing power for the enterprise
operating system of multimodal AI.
The
regulatory dimension is equally critical: in February 2025, Microsoft announced
the completion of the EU Data Boundary for Microsoft Cloud, designed to store
and process commercial and public sector customer data and pseudonymized
personal data within the European Union and the European Free Trade Association
for core cloud services, including Microsoft 365, Dynamics 365, Power Platform,
and most Azure services (Microsoft, 2025b).
This move
should be read as a response to European pressure for digital sovereignty,
privacy, security, and public control over critical infrastructures, given
that, as Walden and Michels (2022, p. 36) note, “cloud providers who offer
services in multiple Member States can be subject to the concurrent
jurisdiction of several national regulators.” It also shows, with complete
transparency, that the enterprise cloud is not neutral: it defines who can
access data, which rules apply, how incidents are addressed, what contractual
obligations exist, and what room for maneuver States and organizations retain
vis-à-vis global providers.
In fact,
Azure operates through cloud regions: sets of geographically distributed data
centers and network infrastructure designed to reduce latency, strengthen
operational continuity, and comply with data residency rules. Those centers are
Azure’s “territorial dimension” and one of its sensitive points: Microsoft
states that it is redesigning its data centers for AI and cloud with
technologies such as direct-to-chip cooling and hybrid wood-steel construction
in order to reduce water consumption and embodied carbon (Microsoft, 2026, p.
16).
The
geopolitical component, then, appears at the material base of AI, where
companies using Azure contract computing capacity and infrastructure and use
libraries and programs, in a “sale of cloud services” that Azure uses to
become, together with Amazon and Google, one of the major organizers of the
material conditions that make AI possible.
Within this
framework, Microsoft is a powerful actor of infrastructural power: it does not
impose public norms, but it creates technical and contractual conditions that
guide behavior.
The paradox
is that Azure can enable safer, more scalable, and more governable AI adoption
for organizations that could not build that infrastructure on their own, but at
the cost of deepening their dependence on a private platform. AI governance
becomes more possible, but also more concentrated.
In
conclusion, Microsoft Azure is one of the very few spaces of power where the
following are established:
- Infrastructure, through the
concentration of computing power, storage, networks, data centers, and
specialized capacities for training, inference, and deployment of AI
systems.
- Integrations, through the
connection of models with Microsoft 365, Copilot, GitHub, Dynamics, Power
Platform, digital identity, security, and corporate data.
- De facto regulations, by
defining terms of use, security policies, evaluation mechanisms,
responsible AI tools, content controls, and practical governance
standards.
- Massive profits, through value
capture from cloud consumption, licenses, managed services, support,
integration, data, productivity, and dependence on the Microsoft
ecosystem.
While
Azure’s public narrative rests on efficiency, innovation, security, and
productivity, its position as an actor of power in the field of AI shows it as
a private infrastructure that organizes public and private capacities to adopt,
control, scale, or limit intelligent systems, and to capture the value produced
by those who host their models and products there, reinforcing patterns of
infrastructural dependence among its users.
Basic data
- Azure is part of Microsoft’s enterprise ecosystem and functions as a central piece of its cloud, data, cybersecurity, and artificial intelligence strategy. Its control lies with Microsoft’s cloud and artificial intelligence division and with the company’s CEO: Satya Nadella.
- Microsoft Foundry turns Azure into a “model marketplace”: it provides access to models from Microsoft, Azure OpenAI, Anthropic, Meta, Mistral AI, DeepSeek, and other providers.
- Azure China operates as a separate cloud through 21Vianet, not as an ordinary part of the global public cloud.
- Azure does not have an independent market valuation, because it is part of Microsoft Corporation, but its economic scale is enormous: in fiscal year 2025, it surpassed US$75 billion in annual revenue, with year-on-year growth close to 34%.
- Microsoft reported Microsoft Cloud revenues of 168.9 billion dollars in fiscal year 2025, with Azure as the central piece around which other products are grouped.
- Azure competes directly with AWS and Google Cloud, but its distinctive advantage lies in its already installed enterprise ecosystem.
References
Microsoft.
(2025a). Microsoft 2025 annual report. https://www.microsoft.com/investor/reports/ar25/index.html
Microsoft. (2025b,
February 26). Microsoft completes landmark EU Data Boundary, offering enhanced
data residency and transparency. Microsoft On the Issues. https://blogs.microsoft.com/on-the-issues/2025/02/26/microsoft-completes-landmark-eu-data-boundary-offering-enhanced-data-residency-and-transparency/
Microsoft.
(2026). Microsoft 2026 Environmental Sustainability Report: Responsibly
building the AI future. Reporting on our 2025 fiscal year. https://cdn-dynmedia-1.microsoft.com/is/content/microsoftcorp/microsoft/msc/documents/presentations/CSR/2026-Microsoft-Environmental-Sustainability-Report-PDF.pdf
Microsoft
Learn. (2026). Microsoft Foundry Models overview. https://learn.microsoft.com/en-us/azure/foundry/concepts/foundry-models-overview
Microsoft.
(2026a, April 27). The next phase of the Microsoft-OpenAI partnership. The
Official Microsoft Blog. https://blogs.microsoft.com/blog/2026/04/27/the-next-phase-of-the-microsoft-openai-partnership/
Microsoft
Azure. (2026). Foundry Models. https://azure.microsoft.com/en-us/products/ai-foundry/models/
Walden, I.,
& Michels, J. D. (2022). Getting critical: Making sense of the EU
cybersecurity framework for cloud providers. Computer Law & Security
Review, 45, 105701.