Meta becomes one of Microsoft’s largest AI customers

Meta Platforms has emerged as one of Microsoft’s largest artificial intelligence customers which highlights the growing demand for computing power and AI models even among companies investing billions of dollars in their own AI infrastructure.

According to a Bloomberg report published on Thursday, Meta is spending hundreds of millions of dollars annually to access AI models through Microsoft’s Azure cloud platform.

The development is notable because Meta has been pursuing an aggressive strategy to build its own AI infrastructure.

The Facebook and Instagram parent has committed substantial resources to data centres, AI chips and computing capacity as it develops its next generation of AI products.

Meta has also been expanding its use of multiple infrastructure partners rather than relying on a single provider. In April, the company announced a partnership with Amazon Web Services to add tens of millions of AWS Graviton cores to its computing portfolio for agentic AI workloads.

Meta’s long-standing Azure relationship

Meta’s relationship with Microsoft Azure is not new. In 2022, Microsoft announced that Meta had selected Azure as a strategic cloud provider to accelerate AI research and development.

Under that arrangement, Meta expanded its use of Azure’s supercomputing capabilities, including a dedicated cluster containing 5,400 NVIDIA A100 GPUs for some of its large-scale AI research workloads. Meta had already begun using Azure virtual machines for AI research in 2021.

The latest development suggests that the relationship has grown significantly as the computing requirements of modern AI systems have expanded.

Why Meta still needs Microsoft

Meta has been spending heavily to build its own AI computing capacity, but developing advanced AI systems requires enormous amounts of computing power.

The company has been investing in custom AI silicon through a partnership with Broadcom and has been expanding its data-centre footprint. In July, Meta also announced a venture with BlackRock and its infrastructure investment partners to develop a data-centre campus in El Paso, Texas.

Meta’s infrastructure push came with significant financial costs. Reuters reported in July that Meta’s free cash flow fell sharply in the second quarter as the company increased spending on chips, servers, energy and data centres to support its AI ambitions.

Using Microsoft’s Azure infrastructure therefore gives Meta additional computing capacity while it continues building out its own systems.

A significant boost for Microsoft’s AI business

For Microsoft, Meta’s spending demonstrates the commercial opportunity created by Azure’s AI infrastructure.

Microsoft has positioned Azure as a major platform for companies that need access to advanced AI models and large-scale computing without building all of the underlying infrastructure themselves.

Meta’s reported spending also illustrates an important feature of the AI market that even the world’s largest technology companies are becoming customers of one another as the cost of developing and running AI systems rises.

This development comes as Meta itself considers turning some of its growing computing infrastructure into a commercial business.

The company was exploring plans to sell AI computing capacity and models to external customers, potentially putting it in competition with Microsoft Azure, Amazon Web Services and Google Cloud.

For Meta, the Azure relationship shows the scale of its AI ambitions while for Microsoft, it reinforces Azure’s position as an infrastructure layer not only for traditional enterprises but also for some of the world’s biggest AI developers.