To understand why this matters, you need to know what the old deal looked like. Back in 2019, Microsoft started investing in OpenAI, eventually putting in over $13 billion across several rounds. In exchange, Microsoft got something extremely valuable: exclusive access to OpenAI's models and technology, and the right to be the only cloud provider allowed to run them. Every time someone used ChatGPT or the OpenAI API, it ran on Microsoft's Azure servers. Microsoft also received a revenue share, meaning they got a cut of OpenAI's earnings. It was one of the most consequential partnerships in tech history, and it made Microsoft look like the early winner of the AI race.

Today that deal got quietly dismantled. Microsoft's license to OpenAI's models and technology will no longer be exclusive, meaning OpenAI can now offer its products and services through any cloud provider, including Amazon and Google. Microsoft will no longer pay a revenue share to OpenAI, though OpenAI will continue paying Microsoft a 20% cut of its own revenue through 2030, subject to a total cap. Microsoft remains OpenAI's primary cloud partner and gets products first on Azure, but the exclusivity that made the original deal so powerful is gone.

OpenAI has been striking deals with Microsoft's competitors for months, including a major partnership with Amazon Web Services. The new agreement formalizes what was already happening in practice. OpenAI has outgrown the original arrangement, and Microsoft had to choose between renegotiating on OpenAI's terms or losing the relationship entirely. They renegotiated.

The headline reads like a breakup. The reality is more like a renegotiation of power. OpenAI needed the freedom to grow. Microsoft kept the access, the revenue stream, and the equity stake. Whether that trade was worth it is the open question.

OpenAI was founded in 2015 as a nonprofit explicitly built to develop AI for the benefit of humanity, free from shareholder pressure and profit motives. Elon Musk was a co-founder and contributed around $38 million in seed funding. The founding documents said the technology must benefit all of humanity. That was the whole point.

Then OpenAI raised billions, built ChatGPT, hit a valuation approaching $1 trillion, and quietly restructured into a for-profit company. Musk, who had left the board in 2018, filed suit in 2024 arguing that what happened between 2015 and today is a straightforward betrayal of the founding promise. Two claims remain going into trial: breach of charitable trust, meaning he is arguing that OpenAI took assets donated for a public good and converted them for private gain, and unjust enrichment, meaning Sam Altman and others personally benefited from something they were never entitled to profit from.

He is seeking up to $134 billion in damages, though he has said any winnings would go back to the nonprofit to restore its original mission, not into his own pocket. That detail matters because OpenAI has been framing the lawsuit as a competitive attack from a bitter rival.

OpenAI's response is that Musk himself was involved in early discussions about creating a for-profit structure before he left the board, and that the whole thing is a tactic to damage a competitor while he prepares xAI for its own IPO. They posted online that they cannot wait to question him under oath. Musk posted that Altman stole a charity. Prediction markets have the odds at roughly 50-50.

Musk, Altman, and Microsoft CEO Satya Nadella are all named as witnesses. The liability phase runs through mid-May. The emails coming into evidence are going to be worth reading.

Symphony is a new open-source tool that connects Codex to your task tracker, turning every open issue into an agent task that runs automatically in the background. Humans stay in the loop for review and direction, agents handle the rest. Someone described it as Notion for AI agents, which is probably the most accurate one-line summary. Worth keeping an eye on.