Microsoft Invests US$1 Billion in OpenAI

Microsoft first invested in OpenAI on 22 July 2019. That day the two companies, announcing from San Francisco and Redmond, set out a multiyear partnership in which Microsoft invested US$1 billion in OpenAI. The partnership had three parts: the companies would jointly build new Azure AI supercomputing technologies; OpenAI would port its services to run on Azure; and Microsoft would become OpenAI's preferred partner for commercialising new AI technologies.

At the time OpenAI was a research organisation with no ChatGPT and not even an API product. This page checks what the two companies actually announced that day, against their own words.

What was announced on 22 July 2019

The wording differed. Microsoft's headline spoke of an "exclusive computing partnership", and its body said OpenAI would port its services to Azure. OpenAI's own post was blunter: Microsoft would become its exclusive cloud provider.

Microsoft's release carried a line from OpenAI's CEO, Sam Altman: "The creation of AGI will be the most important technological development in human history, with the potential to shape the trajectory of humanity."

Why OpenAI needed the money

OpenAI began in 2015 as a non-profit research lab. On 11 March 2019 it announced OpenAI LP, a "capped-profit" company attached to the non-profit, explaining that it would need to invest billions of dollars in the coming years in large-scale cloud compute, talent and AI supercomputers. Investors and employees could earn a capped return; anything beyond that would belong to the original non-profit. Microsoft's investment came four months after the new structure was set up.

OpenAI's post of 22 July also explained how it meant to pay its way. Building a product was the obvious route, but that would mean changing its focus. Instead it would license some of its pre-AGI technologies, with Microsoft as the preferred partner for commercialising them.

On the technical side, OpenAI had announced the GPT-2 language model on 14 February that year (the background is in the GPT-3 article). The July post listed a step in AI capability for each year since 2012, with writing text as the example for 2019.

How the partnership grew

On the 2020 GPT-3 licence, Microsoft CTO Kevin Scott wrote that OpenAI would continue to offer GPT-3 and other models through its own Azure-hosted API. What Microsoft held exclusively was the right to build the model into its own products.

In January 2023 Microsoft called the new deal the third phase of the partnership, following "our previous investments in 2019 and 2021". It described the investment only as "multiyear, multibillion dollar", without a figure, and reaffirmed Azure as OpenAI's exclusive cloud provider, powering all its research, products and API services.

How exclusivity loosened

On 21 January 2025 Microsoft said that OpenAI had made a new, large Azure commitment, and that exclusivity on new capacity was moving to a model in which Microsoft has a right of first refusal. Microsoft also approved OpenAI building additional capacity of its own, primarily for research and model training.

The agreement of 28 October 2025 accompanied OpenAI's conversion of its for-profit arm into a public benefit corporation, OpenAI Group PBC. Microsoft said its stake was worth about US$135 billion, roughly 27 percent on an as-converted diluted basis including all owners. At that point Microsoft kept exclusive IP rights and Azure API exclusivity until AGI.

Then, on 27 April 2026, the two companies revised the deal again. According to The Next Web, Microsoft keeps a licence to OpenAI's intellectual property through 2032, but it is now non-exclusive; Azure remains the primary cloud, while OpenAI may serve its products through other cloud providers. The exclusive relationship begun in 2019 changed shape substantially.

Where the 2019 deal stands

Measured against what followed, US$1 billion was a modest sum. But the July 2019 announcement fixed the frame of the partnership: OpenAI's compute on Azure, and Microsoft as the channel for commercialising its technology. The 2023 investment, the new Bing, and OpenAI models inside Microsoft's products were all built on that frame.

Microsoft's CEO at the time was Satya Nadella, who had made the cloud the centre of the company since taking over in 2014 (see his appointment). For the companies see Microsoft and OpenAI; for the wider story, the AI timeline.