OpenAI is ending Cursor’s direct access to its AI models after Elon Musk’s SpaceX acquired its parent company, Anysphere. This is the next move in the increasingly bitter conflict between Elon Musk and the company he helped found.

Just because you’re filthy rich doesn’t mean you still can’t have petty fights. In the latest example of big-wig tech fusses, OpenAI said it notified SpaceX on Aug. 28 that it intends to wind down its Cursor contract with a Nov. 12, 2026 shutoff date. Why? OpenAI said, “We are making this choice because we cannot be confident that SpaceX will use our technology within our terms of service.“

After all, OpenAI continued, “After Musk acquired Twitter, now part of SpaceX, the company broke⁠ the terms of our contract (alongside many others). Under oath earlier this year, Musk admitted⁠ that xAI, now also part of SpaceX, had violated OpenAI’s terms of service (terms which are similar to xAI’s own).”

So, OpenAI is invoking a limited termination right in its Cursor agreement following SpaceX buying Cursor to end the deal. OpenAI is declining to provide new OpenAI models to the coding tool during the transition.

Elon Musk replied on X, “I couldn’t care less. Scam Altman and Greg Stockman are utterly untrustworthy assholes who stole an open-source nonprofit.” Yep, there’s no love lost here.

The decision may or may not hurt Cursor too much. The AI programming company also offers models including Claude Fable, Opus, and Sonnet; Gemini 3.1 Pro and Gemini 3.7 Flash; and SpaceXAI’s own Grok LLMs. Still, it cuts one of the most prominent AI coding environments off from a longtime foundation-model partner and underscores how access to leading AI models has become leverage in an increasingly competitive developer-tools business.

OpenAI said it had worked with Cursor for nearly four years and described the move as “incredibly tough,” noting that developers who use OpenAI models through Cursor would be the people most directly affected. The company said it will keep models available until the latest date permitted under the agreement, but will withhold future releases, including its upcoming Astra model, from Cursor.

Cursor CEO Michael Truell tried to make light of OpenAI’s move. He said on X that OpenAI models only serve about 5% of the platform’s user traffic. Truell added, “Cursor was one of the very first users of OpenAI, we’ve worked closely with their team for years, and we’ve trusted their platform to be neutral infrastructure for our business.”

On X and other social networks, most people found his claim that only 5% of Cursor work went through OpenAI models to be laughable.

The cutoff is a potentially significant, but hardly existential, disruption for Cursor. In the meantime, Tom Brown, Anthropic founder and Chief Compute Officer, tweeted, “Cursor has been a trusted partner of Anthropic since Sonnet 3.5. We’ll continue to increase compute to support Claude models in Cursor.” Still, developers whose work relies on particular OpenAI model behavior, coding workflows, or account arrangements will need to migrate or reconfigure before the proposed November deadline.

The Cursor split is the latest extension of the long-running cat-fight between Musk, OpenAI Chief Executive Sam Altman and OpenAI President Greg Brockman. Musk co-founded OpenAI in 2015 as a nonprofit AI research laboratory and was one of its early financial backers. He left the company’s board in 2018 after disagreements over its direction. Musk followed this by hiring away talent and cutting off promised donations.

Since then, OpenAI has become a commercial AI heavyweight through ChatGPT, its Microsoft partnership and a corporate structure combining a for-profit business with nonprofit oversight. Musk has argued that OpenAI departed from its original commitments to develop advanced AI for the public benefit rather than private gain. OpenAI says that’s not the case.

Musk sued OpenAI, Altman and Brockman in 2024 over the company’s restructuring. While he lost that lawsuit earlier this year, he’s said he’ll appeal. Meanwhile, Altman and Brockman claim that Musk knew all along that the plan was to move OpenAI from being a non-profit organization to a for-profit company. They added, “We’re sad that it’s come to this with someone whom we’ve deeply admired—someone who inspired us to aim higher, then told us we would fail, started a competitor, and then sued us when we started making meaningful progress towards OpenAI’s mission without him.”

The bigger issue than AI fat cats feuding with each other is that large integrations such as the one with Cursor generally require custom contracts to establish safety and terms-of-service compliance at scale. As cross-platform agentic work becomes more commonplace, there will be many more such deals, and they’ll also be vulnerable to disruption when AI businesses fight each other.

This is a reminder that ostensibly model-neutral AI coding tools can be vulnerable when a major supplier decides a new owner has become an unacceptable rival. This isn’t the first time this has happened, for example, in June 2025. Anthropic blocked programming AI company Windsurf’s access to Claude models after it emerged that OpenAI would be buying Windsurf. That deal fell apart, but it set a precedent that OpenAI is now using.

So, yes, this is just another battle in the Musk vs. OpenAI fight. It’s also, however, a worrying example of how AI corporate fights can interfere with users’ AI work. I fear this will be only the first of many other such conflicts.