OpenAI frames the break as a contractual and safety safeguard; Cursor sees a threat to neutral infrastructure it has relied on for years. SpaceX and Musk, meanwhile, have signaled that Anthropic’s Claude models could soften the blow.

The dispute follows SpaceX’s $60 billion acquisition of AI coding startup Cursor on Aug. 14, folding the company into Musk’s widening technology empire. Cursor CEO Michael Truell said OpenAI models account for roughly 5% of user traffic and that the company is talking with OpenAI to resolve the issue. “We’ve trusted their platform to be neutral infrastructure for our business,” he said.¹

Late Friday, OpenAI notified SpaceX that it intends to wind down its Cursor contract, setting Nov. 12, 2026 as the proposed shutoff date—the latest cancellation point permitted under the agreement. The company said it would not supply future models during the transition.²

OpenAI’s rationale reaches beyond Cursor’s product. It said the change in ownership triggered a limited cancellation window and argued it could not be confident SpaceX would follow its terms, citing what it says were prior contract breaches by Musk-owned Twitter and xAI. With its forthcoming Astra model in view, OpenAI said it needs stronger compliance and safety controls, while promising support for developers caught in the middle.²

Musk’s response was not a rebuttal of the contract allegations so much as a signal of an alternative. He posted rocket emojis while sharing Anthropic’s pledge to increase computing capacity for Claude inside Cursor after the SpaceX deal.Elon Musk (elonmusk)³ Anthropic’s support gives Cursor a route around OpenAI; it does not resolve the deeper feud between Musk and the company he helped found.