Microsoft is trying to define how it will handle access to major AI infrastructure at a moment when its role in the generative AI market is under close watch. At Mobile World Congress in Barcelona, the company introduced an eleven-point framework called AI Access Principles, aimed at explaining how it plans to operate AI datacenter infrastructure and other AI assets around the world.
The announcement follows Microsoft’s investment and partnership with French Large Language Model startup Mistral AI, and arrives amid continuing questions about whether its deep relationship with OpenAI could limit competition.
What Microsoft Put On The Table
The AI Access Principles are framed as a broad guide for how third parties can use Microsoft platforms and services to build AI products. That matters because Microsoft sees this as a critical business area, not only for carriers attending MWC, but also for businesses and organizations across many industries.
The framework covers several areas. Microsoft said it will support an app store approach that lets businesses choose between different LLMs and other AI products. It also said customer proprietary data will not be placed into Microsoft training models.
Another commitment is cloud flexibility. Microsoft said customers should be able to change cloud providers, or change services within the cloud, if they decide to do so.
- Support for multiple LLMs and AI products through an app store model.
- A commitment not to use company proprietary data in training models.
- The ability for customers to change cloud providers or cloud services.
- Cybersecurity work around AI services.
- Attention to environmentally-sound data centers and infrastructure.
- Education investments connected to AI access.
The OpenAI Question Behind The Announcement
The timing is central to the story. Microsoft is facing increasing regulatory scrutiny over its $13 billion investment in OpenAI, which currently gives it a 49% stake in the startup. OpenAI is described in the source as leading the charge for generative AI services globally.
In January, the European competition watchdog said it was assessing whether the investment falls under antitrust rules. Against that background, Microsoft’s principles read as both a public policy statement and a business reassurance campaign.
Brad Smith, the president and vice chair of Microsoft, announced the framework at Mobile World Congress. The company’s message was one of openness to stakeholders, though the announcement came through a keynote speech with no scope for follow-up questions.
Data Use Is The Core Promise
One of the most direct parts of Microsoft’s message concerns customer data. The company is trying to reassure businesses that using Microsoft infrastructure for AI development will not turn their own data into a competitive weapon against them.
“If they are training a model on our infrastructure, if they are deploying it on our infrastructure, we recognise that their data is their data, we will not access it and use it to compete with the companies that are relying on our infrastructure,” Smith said.
For companies building AI systems, that statement addresses a practical concern. If an enterprise trains or deploys a model on another company’s infrastructure, it needs confidence that its proprietary information will remain separate from the platform provider’s own competitive interests.
Microsoft also used model choice as part of its argument. Smith said on stage that, “In fact, as of today, we have almost 1600 models running in our data centres, 1500 of which are open source models,” presenting that mix as evidence that the company works with proprietary and open source models, and with companies both large and small.
Why Non-Binding Principles Still Matter
The AI Access Principles are not binding rules for Microsoft. The source also notes that Microsoft did not provide detail on how the commitments would be verified or tracked. That limits what the framework can prove on its own.
Still, a public framework can have consequences. If formal regulatory investigations develop, Microsoft will likely point to the principles as evidence that it is taking proactive steps to protect competition in AI markets.
The same public nature can also work the other way. Competitors, customers, the public and regulators can use the AI Access Principles as a reference point if they believe Microsoft is not living up to what it announced.
That makes this more than a branding exercise. It is a public record of how Microsoft says it will handle AI infrastructure, customer data, model choice, cloud mobility, cybersecurity, sustainability and education as its AI business grows under regulatory attention.