Software King of the World, Microsoft, is weighing a legal punch-up with Amazon and OpenAI over a $50bn deal that could blow a hole in its exclusive cloud arrangement with the ChatGPT outfit.
The row is about whether Amazon Web Services can sell OpenAI’s new commercial product, Frontier, without breaking a contract that routes model access through Microsoft’s Azure.
That deal has been a nice little earner for Vole, with OpenAI products helping push Azure revenues to record highs.
Amazon and OpenAI claim they are building a workaround, but Microsoft executives say it is fantasy tech and a breach in spirit if not in ink.
Talks were still going on ahead of Frontier’s launch, with everyone trying to dodge court and keep the mess behind closed doors.
According to a Volish deep throat who spoke to the FT: “We know our contract. We will sue them if they breach it. If Amazon and OpenAI want to take a bet on the creativity of their contractual lawyers, I would back us, not them.”
It underlines a wider split, with OpenAI wanting to wriggle out of early constraints and spread its cloud bets as Microsoft starts treating it like a rival in enterprise AI.
OpenAI insists its Amazon plans fit the Microsoft deal, according to someone familiar with its thinking.
That person said Microsoft is unlikely to sue while it faces regulatory probes in the US, UK and EU into alleged anti-competitive licensing with Azure.
OpenAI’s hopes for a public listing as early as 2026 could be snarled if the dispute lands in court, even after a $110bn funding round last month.
It still needs more cash to pay for the compute burn of training and running large language models, which is not the sort of tab you pick up with a shrug.
The IPO picture is already messy thanks to a lawsuit from Elon Musk against OpenAI, chief executive Sam Altman, which is due to go to trial in Oakland in April 2026.
Musk accuses Altman of ditching OpenAI’s non-profit mission to enrich himself and other executives, and that fight is not going away quietly.
One person close to Microsoft’s view put it like this: “The last thing OpenAI needs is another court case right now.”
Frontier is pitched as fleets of AI agents running inside companies, with bots operating under human instructions, and it is the centrepiece of the OpenAI-Amazon tie-up announced last month.
That partnership came with a pledge to buy $138bn in cloud services from AWS, which tends to focus minds in any boardroom.
Microsoft had been OpenAI’s exclusive cloud provider since putting $1bn into the start-up in 2019, then gave up that exclusivity when it signed off on a restructuring in October.
It kept a clause covering application programming interfaces, the pipes developers use to access OpenAI models, requiring all API calls to be routed through Azure.
Lawyers have been squabbling for weeks over what Amazon’s agreement really covers and how the companies can describe it without tripping wires.
To avoid poking Microsoft in the eye, Amazon has told staff to be careful describing SRE, according to an internal memo seen by the Financial Times and first reported by Business Insider.
AWS staff can say SRE is “powered by”, “enabled by” or “integrates with” OpenAI, but they are banned from saying it “enables access” or “calls on” ChatGPT.
They are told not to suggest OpenAI’s most advanced frontier models are available on AWS, which is a telling line for something meant to be shouted from the rooftops.
