Elon Musk has launched another blistering attack on OpenAI’s leadership after the AI company announced plans to withdraw its models from Cursor following the coding platform’s acquisition by SpaceX.
The SpaceX chief described OpenAI CEO Sam Altman and President Greg Brockman as “utterly untrustworthy assholes,” reviving his long-running accusation that the organisation abandoned the nonprofit principles on which it was originally founded.
Musk was among OpenAI’s co-founders in 2015 alongside Altman, Brockman and others, but departed in 2018 following disagreements over the organisation’s direction. Their relationship has since deteriorated into a highly public dispute involving lawsuits, competing AI ventures and accusations over OpenAI’s transformation into a major commercial enterprise.
The latest confrontation follows SpaceX’s reported $60 billion acquisition of Anysphere, the company behind AI coding platform Cursor. Cursor allows developers to work with artificial intelligence models supplied by several competing providers.
OpenAI announced on Friday that it intends to terminate Cursor’s access to its models, proposing November 12 as the cutoff date.
The company said the decision followed concerns about whether SpaceX would comply with its contractual requirements, stating it could not be confident that “SpaceX will use our technology within our terms of service” given its previous dealings with Musk’s companies.
Musk responded on X with a characteristically personal broadside.
“I couldn’t care less. Scam Altman and Greg Stockman are utterly untrustworthy assholes who stole an open source nonprofit,” he wrote.
OpenAI’s models currently represent only about 5% of Cursor’s user traffic, according to Anysphere co-founder Michael Truell. He said discussions with OpenAI were continuing in an effort to settle the disagreement.
Truell also pointed out that Cursor had been an early OpenAI customer and said his company had relied on OpenAI’s platform functioning as “neutral infrastructure” on which businesses could build.
OpenAI maintains that its contract with Cursor permits termination within a specified period following a change in ownership. The company has not alleged that Cursor itself breached the agreement.
Rival AI developer Anthropic has meanwhile moved quickly to exploit the opening, announcing that it would expand computing capacity for its Claude models used through Cursor.
The dispute adds another chapter to Musk’s prolonged battle with the company he helped establish. Musk reportedly contributed around $38 million during OpenAI’s early years and sued the organisation in 2024, arguing that he had been induced to support the venture on the understanding that it would remain nonprofit.
OpenAI has subsequently undergone major restructuring and is now reportedly valued at approximately $850 billion. Its nonprofit parent retains control and an ownership interest in the commercial operation responsible for ChatGPT.
Musk has argued that OpenAI betrayed its original mission by embracing a profit-oriented structure and accepting Microsoft’s multibillion-dollar investment. He has repeatedly accused Altman of effectively having “stolen a charity” and has called himself a “fool” for helping finance the organisation.
A federal jury rejected Musk’s claims in May after finding that his action had been brought outside the applicable statutory deadline. Musk has said he intends to appeal.
Image credit: Andrew Neel
