A Look at the Unraveling Microsoft/OpenAI Partnership -- Redmondmag.com

Generation AI

A Look at the Unraveling Microsoft/OpenAI Partnership

By now you have undoubtedly seen reports of friction between the Microsoft and OpenAI camps. What had been a very lucrative and beneficial partnership seems to be slowly imploding, and while both sides are being diplomatically quiet, the signs of trouble are there. So how does it all end? Let's review.

The alliance between Microsoft and OpenAI, creator of the ChatGPT service, dates back to 2019 when Microsoft invested $1B in OpenAI to provide exclusive supercomputing functions to the Azure service. The alliance was furthered in January 2023 with a $10 billion investment on Microsoft's part into OpenAI.

The first signs of a problem began to emerge in June 2023, when Microsoft rushed GPT-4 integration into Bing and OpenAI warned against premature launch.

But the real drama began that November, when the OpenAI board ousted CEO Sam Altman, completely blindsiding Microsoft. Microsoft quickly moved to hire Altman but the OpenAI board backtracked and Altman was reinstated. Still, this had to put a lot of doubt in Microsoft’s collective mind.

In March 2024, Microsoft acquired most of Inflection AI’s team and hired its founder and CEOo Mustafa Suleyman, who is also the founder of DeepMind, an AI service acquired by Google. He now serves as CEO of Microsoft AI. This sent a clear signal that Microsoft intended to build in-house AI.

OpenAI then starts advocating for multi-cloud access, looking at Google Cloud and Oracle Cloud Infrastructure. In June 2024, OpenAI cut a $10 billion deal with Oracle for extra compute capacity. Later in the year, OpenAI would partner with SoftBank, Thrive, Nvidia and Oracle on the Stargate AI data center project, which notably omits Microsoft.

Throwing more fuel on the fire was when OpenAI started going directly to customers and encouraging the customers to deal directly with them, knocking out Microsoft as the middleman. Needless to say, this didn't sit well with Microsoft.

But it could be argued that Microsoft brought this on itself with the hiring of Suleyman, a significant figure in the AI space, to lead Microsoft AI efforts. That sent the signal that it was eventually going to cut OpenAI loose, so OpenAI responded by going into business for itself.

However, Microsoft moves should come as no surprise. The sudden firing and then rehiring of Altman gave the impression of instability at OpenAI And no one wants an unstable business partner. Plus, Microsoft does not have a history of partnering; it goes it alone and develops its own products rather than partnering and licensing. So OpenAI's days were always going to be numbered.

Microsoft does not have a track record of the partnerships. It has a track record of building its own. Go back to the ‘90s when it licensed the Spyglass web browser to compete with Netscape. Very quickly it came out with Internet Explorer. Microsoft goes its own way and it wasn't going to be any different with OpenAI. The partnership with OpenAI was merely stopgap on the way to doing its own thing.

I am not privy to the language of the contracts between the two companies, so I cannot say what's binding and what might be in breach of contract. But given that OpenAI is already working with the Stargate project and Microsoft has said nothing (at least publicly) there either isn't a non-compete clause or Microsoft just doesn't care at this point.

So this will likely end with Microsoft owning its own technology and OpenAI cut loose. I think Microsoft recognizes that it has significantly greater resources than any small AI company and once they get a performance technology competitive with ChataGPT, they're going to want to run with it with their resources behind it in a way they see fit.

Posted by Andy Patrizio on 07/11/2025

Featured

Where Air Gapped Backups Actually Fail, Part 2

Air gapped backups can still fail when configuration drift, lost encryption keys and routine human mistakes go unnoticed until recovery is needed.
Nadella Says Enterprise AI's Future May Depend Less on Frontier Models Than Learning Systems

Satya Nadella is making the case that the next phase of enterprise AI will not be won by the company with the highest-scoring model on a public leaderboard.
AI Agents Could Put Enterprise GenAI Budgets Under Pressure

Enterprises moving generative AI from pilots into production may be underestimating the cost of running those systems at scale, particularly as inference workloads and AI agents become more common, according to a new Gartner report.
Banks Targeted by Fileless Phantom Stealer Phishing Campaign

A new phishing campaign is targeting banks and other high-value organizations with Phantom Stealer, a commercially available infostealer that runs in memory to avoid traditional detection, according to new research from Fortra.
Where Air Gapped Backups Actually Fail, Part 1

Air gapped backups can provide critical ransomware resilience, but restore testing gaps, documentation drift and media rotation mistakes can leave organizations unsure whether they can actually recover.