Court documents released this week give a detailed account of the early days of Microsoft’s partnership with OpenAI. When OpenAI was developing AI gaming bots, Microsoft CEO Satya Nadella and OpenAI CEO Sam Altman were starting their collaboration. The documents, from the Musk v. Altman trial, reveal concerns that the AI startup might partner with Amazon instead and criticize Microsoft. In 2017, after OpenAI’s bot defeated a Dota 2 pro player, Altman proposed a major partnership to Nadella, seeking $300 million beyond Azure credits. This proposal worried some Microsoft executives about direct revenue. Jason Zander, then Azure chief, expressed the concern that $500 million in revenue was necessary to justify such a deal. Altman later suggested a partnership involving Xbox and sharing AI technology in exchange for supporting OpenAI’s Dota research. However, the Xbox team couldn’t afford the research costs alone. In early 2018, Microsoft CTO Kevin Scott expressed concerns about OpenAI switching to Microsoft’s cloud rival. He highlighted the potential PR damage if OpenAI left for Amazon. He acknowledged OpenAI’s growing influence in AI but questioned the value of their demands. By 2019, Scott admitted his initial skepticism about AI from OpenAI and DeepMind shifted when OpenAI progressed to natural language processing models. Concerned about Google’s advancement in AI, Microsoft invested $1 billion in OpenAI. Eventually, the partnership evolved, with OpenAI renegotiating the deal to use AWS for bringing its AI models like Codex to enterprises, pointing out the limitations caused by their exclusive arrangement with Microsoft.
Microsoft Worried OpenAI Might Shift to Amazon and Criticize Azure
2 Min Read