OpenAI to Cut Off Cursor's API Access Nov. 12 After SpaceX Acquisition

OpenAI is terminating its contract with AI coding tool Cursor, effective November 12, 2026, after SpaceX acquired the company. OpenAI says the move stems from Elon Musk's track record of breaking contractual agreements.

According to OpenAI, the November 12 date represents the maximum notice period allowed under its contract with Cursor, which includes a clause permitting termination within a limited window following a change in company ownership. SpaceX's acquisition of Cursor triggered that clause.

"It boils down to trust," OpenAI's Thibault Sottiaux said, announcing the decision. He said OpenAI cannot be confident that SpaceX will comply with OpenAI's terms of service going forward.

Citing Twitter data and Grok distillation

OpenAI points to two specific incidents to justify the decision. First, Musk allegedly broke a licensing agreement with Twitter (now X) that gave OpenAI access to the full tweet data feed for roughly $2 million a year to train ChatGPT. According to OpenAI, when Musk discovered the deal's terms in December 2022 after acquiring Twitter, he judged the price too low and cut off access.

Second, OpenAI cites Musk's own admission during litigation between the two companies that he used outputs from other labs' AI models to train Grok — a practice known as distillation. OpenAI says this violates its terms of service, which prohibit using its outputs to train competing models.

OpenAI is also framing the decision partly as a safety measure. The company says its upcoming Astra model reportedly has advanced cyber capabilities, and that this creates "a new level of accountability" for ensuring partners follow usage terms.

What changes for Cursor users

Users accessing GPT models through Cursor will still be able to connect using their own OpenAI API keys, and OpenAI says it will continue supporting access through its IDE extensions built for Cursor. Sottiaux said the move targets Musk specifically, not the broader AI coding tool ecosystem.

Cursor co-founder Michael Truell responded that OpenAI models account for only about five percent of Cursor's total AI traffic, according to Truell. He said Cursor is in active discussions with OpenAI's team to find a resolution, adding: "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."

Part of a pattern in AI model access disputes

This is not the first time a major AI lab has cut off a coding tool over corporate ties to a rival. In June 2025, Anthropic blocked Windsurf's access to Claude models after reports that OpenAI intended to acquire Windsurf. Two months later, in August 2025, Anthropic revoked OpenAI's own API access after learning OpenAI teams had used Claude internally to benchmark against GPT-5 ahead of its launch.

What this means

This dispute is fundamentally a business and contractual conflict, not a technical one — no new model or capability is involved. It signals that API access to frontier models is increasingly being weaponized along competitive and personal lines between lab leadership, rather than governed purely by usage terms and payment. For developers building on multiple model providers, the incident is a reminder that platform access can be revoked over ownership changes or corporate rivalries with little warning, even when the affected tool represents a small fraction of a provider's traffic, as Cursor's five percent figure suggests. Expect more such disputes as AI labs increasingly compete on infrastructure control rather than just model quality.