Microsoft invests $1 billion in OpenAI and becomes its cloud partner

Background

By 2019 OpenAI was a research organisation working toward artificial general intelligence (AGI), a term for AI that can perform a wide range of intellectual tasks at a human level. Training and running increasingly large AI models needed very large amounts of computing power, which OpenAI could not easily provide on its own.

What happened

On 22 July 2019 OpenAI and Microsoft announced a multiyear partnership. Microsoft invested $1 billion in OpenAI.^[1] The announcement described it as an exclusive computing partnership: OpenAI would port its services to run on Microsoft Azure.^[1]

The two companies planned to build a large-scale Azure computing platform to train and run increasingly advanced AI models. The platform was meant to support OpenAI's work toward AGI.^[1] Microsoft was to become OpenAI's preferred partner for commercialising new AI technologies.^[1]

Key details

OpenAI CEO Sam Altman said "The creation of AGI will be the most important technological development in human history". Microsoft CEO Satya Nadella called AI "one of the most transformative technologies of our time".^[1]

Microsoft later described the 2019 deal as the first phase of the partnership. Further investments followed in 2021 and in January 2023.^[2]

Impact

The 2019 deal set the pattern for what followed. Microsoft became the exclusive computing partner for OpenAI's services, and the partnership was later extended in 2021 and January 2023.^[1][2]

What a compute partnership means

Training a large AI model is a very costly process. It requires many specialised processors running for long periods, and the data centres that hold them. A company that wants to build such models must either buy this capacity or rent it from a cloud provider. An exclusive computing partnership means that one cloud provider is the main place where the model is trained and run. For OpenAI this made Azure its main home for large-scale work, and for Microsoft it meant that Azure's AI infrastructure was built around OpenAI's needs.

Why the deal drew attention

The partnership combined an investment with a supply arrangement. Investors put money in, and the company receiving the money agreed to use the investor's cloud. This pattern made the partnership a reference point in later discussions of how frontier AI labs are financed and how they depend on large cloud providers. The deal also raised questions about how a research organisation with a stated mission toward AGI would balance that mission with commercial partners, questions that the later agreements in this archive returned to.^[1][2]