DooDooLamb News
Musk vs. Altman Evidence Shows What Microsoft Executives Thought of OpenAI
Brief published May 9, 2026 ยท Original source published May 8, 2026
Original reporting by Maxwell Zeff, Paresh Dave at wired.com.
Automated brief. Verify important details at the original source.
Microsoft Emails Reveal Company's 2018 Strategy: "Keep Your AI Enemies Closer"
In a delicious twist of corporate irony, newly surfaced internal emails show that Microsoft executives spent 2018 treating OpenAI like that friend who keeps borrowing money and might steal your girlfriend, but you can't cut them off because what if Amazon gets to them first.
The emails, released as part of the ongoing legal melodrama between Elon Musk and Sam Altman, paint a picture of Microsoft leadership performing the corporate equivalent of nervous laughter. Here was OpenAI, burning through cash faster than a crypto startup in 2021, making grand pronouncements about artificial general intelligence while basically running a very expensive text completion service. Microsoft's brain trust looked at this scrappy nonprofit with its lofty mission and thought, "These people are either going to change the world or faceplant spectacularly, and we absolutely cannot let Jeff Bezos find out which one first."
The correspondence reveals Microsoft executives wrestling with what venture capitalists politely call "founder risk" and what normal people call "what happens when brilliant people start believing their own press releases." OpenAI in 2018 was still pretending to be a research lab focused on humanity's benefit, despite the obvious fact that training increasingly large language models requires the kind of money that only comes with strings attached. It's like watching someone insist they're just casually dating while moving in together and sharing a Netflix account.
What makes these emails particularly entertaining is how they capture Microsoft's strategic thinking in real time. The company was essentially playing a high-stakes game of hot potato with artificial intelligence. They knew OpenAI was going to need serious backing to compete with Google's DeepMind and Facebook's AI Research lab (back when Facebook still pretended to be about connecting people rather than selling ads for questionable supplements). The choice was simple: either write the checks yourself, or watch Amazon Web Services become the infrastructure backbone for your most promising AI competitor. When you put it that way, suddenly a few billion dollars in cloud credits starts looking like a bargain.
The really prescient part of these internal discussions was Microsoft's recognition that OpenAI's nonprofit structure was essentially a legal fiction waiting to be resolved. You can't burn through hundreds of millions in compute costs while maintaining that you're just a humble research organization dedicated to the betterment of humanity. It's like claiming you're just casually interested in fitness while deadlifting a Honda Civic. Something's got to give, and Microsoft executives correctly predicted that OpenAI would eventually need to choose between its idealistic origins and the practical requirements of training models that could actually compete with Google.
By 2018, the writing was already on the wall for anyone who understood the economics of large-scale machine learning. Training state-of-the-art language models isn't like running a philosophy department; it requires industrial-grade computing infrastructure and the kind of sustained capital investment that makes launching satellites look like a weekend hobby. Microsoft's emails show a company that understood this math better than OpenAI's own leadership seemed to at the time.
Fast-forward to today, and Microsoft's early skepticism looks remarkably well-calibrated. OpenAI has indeed transformed from a nonprofit research lab into a for-profit juggernaut valued at over $80 billion, with Microsoft as both its primary investor and increasingly nervous partner. The relationship has evolved from "strategic partnership" to something more resembling a complicated marriage where both parties are constantly checking their prenup. Microsoft gets exclusive access to OpenAI's models for its products, while OpenAI gets the computing power to train increasingly expensive models and the corporate backing to fend off Google's competitive response.
The current legal drama between Musk and Altman adds another layer of absurdity to this corporate love triangle. Musk, who helped found OpenAI before departing in 2018 (coincidentally around the time these Microsoft emails were being written), is now suing the company for abandoning its nonprofit mission. It's the equivalent of your ex-business partner calling you a sellout while simultaneously launching a competing AI company and trying to poach your talent. The emails provide fascinating context for why Musk might have left: even Microsoft's hardened enterprise software executives were nervous about OpenAI's trajectory, so imagine how someone with Musk's particular relationship to risk and control must have felt.
What these emails really reveal is how even sophisticated corporate strategists can struggle to predict the future of transformative technology. Microsoft was simultaneously right to be skeptical of OpenAI's business model and right to invest anyway. Sometimes the best strategy is to hedge your bets and hope your partner figures out how to turn research breakthroughs into sustainable revenue before the money runs out.
The whole situation serves as a masterclass in how technology partnerships evolve when the stakes get high enough. Microsoft thought they were backing a promising AI research lab; instead, they accidentally became the primary investor in their biggest potential competitor since Google. It's the corporate equivalent of offering to help your friend move, only to discover they're relocating to a mansion and expecting you to carry the piano.