OpenAI fires back at Elon Musk with receipts

There is nothing quite as awkward in tech as a former co-founder suing you for abandoning your morals, only for you to pull out receipts showing they endorsed the exact same roadmap years ago.

OpenAI just published a scathing blog post responding to Elon Musk’s recent lawsuit. Musk, who co-founded the firm in 2015 before departing, sued OpenAI and CEO Sam Altman for allegedly ditching their original non-profit mission to benefit humanity in favor of maximizing profits for Microsoft. His complaint even highlighted OpenAI's intricate corporate web of eight sub-entities and questioned whether the current board has the technical chops to evaluate AGI breakthroughs like the rumored Q*.

Instead of issuing a polite legal denial, OpenAI went nuclear. The company published historical internal emails signed by Musk himself, detailing the messy falling out behind closed doors.

The billion-dollar receipts

According to OpenAI, when the organization was founded in late 2015, Greg Brockman and Sam Altman originally targeted a $100 million raise. Musk pushed back via email, arguing that $100 million sounded hopeless and they should announce a $1 billion commitment instead, promising to cover whatever others didn't.

In reality, the non-profit received less than $45 million from Musk, while other donors contributed over $90 million.

By early 2017, the team realized reaching AGI would require billions of dollars per year in compute. When discussions began about creating a for-profit arm to raise capital, Musk demanded majority equity, initial board control, and the CEO title. When the team refused to give one individual absolute power, Musk withheld promised funds, forcing Reid Hoffman to step in and cover payroll.

Musk then pitched merging OpenAI into Tesla to serve as a cash cow, claiming Tesla was the only entity with even a tiny chance of competing against Google. When that bid failed, he walked away in February 2018, pegging OpenAI’s chances of success at zero and stating he would build his own AGI inside Tesla.

OpenAI also tackled Musk’s complaint about GPT-4 being proprietary. The blog included an exchange where chief scientist Ilya Sutskever told Musk that as they got closer to AI, it would make sense to be less open, explaining that the Open in OpenAI meant everyone should benefit from the fruits of AI, not that they had to share the underlying science. Musk’s reply was a single word: Yup.

Why it matters

This dispute pulls back the curtain on the fundamental contradiction at the heart of modern AI development. OpenAI’s public counterpunch is remarkably effective because it uses Musk’s own words to dismantle his moral high ground. We think it makes it painfully obvious that this feud is less about high-minded open-source ethics and far more about control and bruised ego.

For everyday developers and businesses relying on OpenAI’s APIs, the lawsuit changes nothing operationally in the near term. OpenAI made sure to highlight practical humanitarian deployments, such as Albania speeding up EU accession and healthcare groups simplifying consent forms, to prove the tech is actively helping people despite the closed model.

For the broader industry, however, the fallout sets an uncomfortable precedent. Musk's claims about Q* and board competence reveal how murky the technical definition of AGI truly is. As compute costs escalate into the tens of billions, pure non-profit research is functionally extinct, leaving open-source crusaders and commercial tech giants to fight over who really controls the frontier.