OpenAI-Cursor clash widens as Altman-Musk feud heats up

OpenAI is pulling its AI models from Cursor, the coding assistant Elon Musk's SpaceX bought for $60bn earlier this month, in the latest flashpoint of its long-running war with the billionaire.

According to OpenAI, it will reduce the access on 12 November due to the clause in the contract giving OpenAI the right to terminate the agreement in case of "a change of control".

OpenAI stated that it does not see any guarantees from SpaceX to follow its terms of service since OpenAI believes that there is a pattern in the companies owned by Elon Musk violating their contracts.

Musk dismissed the move on X within hours, writing that he "couldn't care less". He went further, branding Altman and OpenAI president Greg Brockman "untrustworthy" and repeating his long-standing claim that they "stole an open source nonprofit" when OpenAI restructured away from its charitable roots.

Michael Truell, one of the founders of Cursor and now an executive at SpaceX, after being acquired by SpaceX, took a more reconciliatory stance on the issue.

Truell claimed that "Cursor was one of the first to use OpenAI" and that it had treated the platform as neutral infrastructure. Meanwhile, Anthropic took the opportunity to expand.

Tom Brown, one of the founders of Anthropic, said that the company will increase its compute to help train Claude models on Cursor, which already has models of Anthropic, Google, and SpaceX's Grok in addition to OpenAI's.

The clash traces back to a courtroom battle in which Musk sought $150bn in damages, accusing Altman of turning OpenAI's nonprofit mission into a vehicle for personal enrichment. A federal jury sided against Musk earlier this year, ruling he had filed too late rather than weighing the substance of his claims.