Microsoft Corp. (NASDAQ:MSFT) has reorganized its Azure cloud leadership and accelerated its custom AI silicon roadmap—scaling its in-house Maia chip to multi-gigawatt deployments by 2027—in a bid to sharpen its focus on artificial intelligence and tighten control over the infrastructure powering products like Copilot and OpenAI models.

Microsoft Reorganizes Azure Around AI

The company reorganized leadership across its Azure cloud business to bring key teams together and sharpen the division’s focus on artificial intelligence.

Cloud chief Scott Guthrie informed employees of the changes on Monday, Oct. 12, the Information reported on Monday.

Arun Ulag, who previously led Azure’s database products, will take on a broader role overseeing the core Azure business and its main engineering teams.

Ulag will also supervise executives running Azure Core, the infrastructure unit that provides computing capacity to major customers including OpenAI.

Maia 200 Moves Into Production

Separately, Microsoft said on Monday that it plans to scale its Azure Maia system to multi-gigawatt deployments starting in 2027.

Maia 200 is already running live Copilot workloads and supports development work for Microsoft’s MAI models.

By year-end, Microsoft expects GitHub Copilot and Microsoft Foundry to run on the system as well.

The company is deploying Maia 200 in the U.S., Europe and Australia and plans to use it across OpenAI models, Microsoft models and major open-source AI systems.

Microsoft said Maia 200 can deliver up to 1.7 times more output per watt than leading GPU infrastructure for AI workloads used in Teams, Copilot and Cowork.

Maia 300 Targets More Efficient AI Computing

Microsoft also plans to deploy Maia 300 in production in 2027.

The company expects the next-generation chip to deliver 40% more AI output per megawatt than leading GPU-based systems.

A full Maia 300 system could produce up to seven times the output of Maia 200 at the rack level, according to Microsoft.

The chip is designed for AI tasks including longer reasoning, faster chat responses and multi-step agent workflows.

Microsoft Builds More Control Over AI Infrastructure

Microsoft describes Maia as a key part of its broader AI strategy, which spans its own chips, cloud systems and software.

Expanding Maia could give Microsoft more control over the infrastructure behind products such as Copilot while supporting larger AI deployments across Azure.

The company plans to provide more details on its strategy, customer adoption and technology roadmap at Microsoft Ignite 2026 next month in San Francisco.

MSFT Price Action: Microsoft shares were up 0.16% at $542.00 during premarket trading on Tuesday. The stock is approaching its 52-week high of $553.72, according to Benzinga Pro data.

Photo via Shutterstock