Microsoft executives had serious doubts about funding OpenAI back in 2017, when the AI lab was still a small nonprofit research operation. Court documents revealed during the Musk v. Altman trial show that Microsoft's leadership team, including CEO Satya Nadella, wavered over what would eventually become one of the most successful corporate partnerships in technology history.

What Made Microsoft Hesitate About OpenAI's Funding Request?

In August 2017, Sam Altman, then OpenAI's CEO, reached out to Nadella after the lab won a video game competition using artificial intelligence. Altman's request was ambitious: he asked for $300 million worth of Microsoft Azure cloud computing services. The problem was that Microsoft's internal teams saw little reason to commit that much money to an unproven startup.

When Nadella asked his team for input, the responses were lukewarm at best. Microsoft's artificial intelligence team saw "no value in engaging" with OpenAI, according to Jason Zander, Microsoft's executive vice president. The research team believed its own work was "more advanced," while the public relations team worried about the optics of supporting a group pushing the idea of "machines beating humans".

A financial analysis made the hesitation concrete: Microsoft calculated it would lose approximately $150 million over several years if it provided the computing services Altman wanted. Zander suggested that while Azure might benefit from associating with Altman and Elon Musk, he wouldn't want to "take a complete bath," or absorb a massive financial loss, in doing so.

Zander

How Did Microsoft's Leadership Debate the OpenAI Decision?

The email chain that unfolded in January 2018 reveals the internal tensions at Microsoft. Nadella forwarded Altman's latest proposal to 15 Microsoft executives, asking for their thoughts. His own assessment was telling: "Overall I can't tell what research they are doing and how if shared with us it could help us get ahead," Nadella wrote.

However, Nadella also acknowledged that Musk was telling people OpenAI was "at the verge of some big AGI breakthroughs," referring to artificial general intelligence, or AGI, which would represent AI systems capable of performing any intellectual task a human can do. Despite this, several executives remained skeptical.

"I'm highly skeptical of an imminent breakthrough in AGI. IMO, they're treating us like a bucket of undifferentiated GPUs, which isn't interesting for us at all," said Kevin Scott, Microsoft's chief technology officer at the time.

Kevin Scott, Chief Technology Officer at Microsoft

Scott raised another concern that would prove prescient: the risk of OpenAI leaving Microsoft for Amazon Web Services (AWS), Amazon's cloud computing division. If that happened, he warned, OpenAI might "bad-mouth us and Azure on the way out." This fear of losing the startup to a competitor seemed to weigh heavily on the decision.

What Changed Microsoft's Mind About OpenAI?

Despite the internal skepticism, Microsoft eventually came around. The company told Altman that no single team would sponsor OpenAI, but the relationship didn't end there. Roughly 18 months after these hesitant emails, Microsoft announced a landmark $1 billion investment in OpenAI after the lab created a for-profit arm. This structure gave Microsoft the potential to generate a return of $20 billion on its investment.

The partnership that Microsoft executives once doubted became extraordinarily valuable. From 2019 through 2023, Microsoft committed $13 billion in cash and cloud computing credits to OpenAI, making it the startup's most prolific fiscal sponsor during that period.

Steps to Understanding Corporate Partnership Risk Assessment

• Financial Modeling: Microsoft calculated potential losses before committing funds, showing how companies quantify risk by projecting costs over multiple years and determining acceptable loss thresholds.
• Competitive Analysis: Executives worried about losing OpenAI to Amazon, illustrating how companies weigh the cost of investment against the risk of competitors gaining advantage.
• Technical Evaluation: Microsoft's teams assessed whether OpenAI's research aligned with their own capabilities and strategic goals, a common practice when evaluating startup partnerships.
• Structural Negotiation: The eventual deal came after OpenAI created a for-profit arm, showing how companies often restructure partnerships to align incentives and reduce risk.

The irony of Microsoft's initial hesitation is striking. The executives who doubted OpenAI's potential were concerned about losing the startup to Amazon. Today, OpenAI has committed to spend $138 billion on Amazon cloud computing services, and Amazon has agreed to invest between $15 billion and $50 billion in OpenAI itself.

These newly released emails were introduced into evidence by Elon Musk's legal team during their lawsuit against OpenAI. Musk's attorneys used the correspondence to show how Microsoft's relationship with OpenAI evolved from skepticism to deep financial entanglement. Musk's lawsuit accuses Microsoft of aiding and abetting unauthorized use of his donations to the AI lab, arguing that Microsoft helped OpenAI develop a moneymaking machine that corrupted its nonprofit principles.

Satya Nadella is scheduled to take the witness stand in the trial on Monday, where he will likely face questions about Microsoft's evolving relationship with OpenAI and the company's role in the startup's transformation from a nonprofit research lab into a for-profit powerhouse.