OpenAI cuts Cursor's model access after its acquisition by SpaceX

[Holly78]

OpenAI announced it's ending Cursor's direct access to its AI models, with a proposed cutoff date of November 12th, following SpaceX's 60 billion dollar acquisition of Cursor's parent company Anysphere earlier this month. OpenAI said plainly that it cannot be confident SpaceX will use its technology within its terms of service, citing what it described as past instances of Elon Musk's companies violating existing contracts, and that it would not provide any future models to Cursor during the wind down period either

Cursor CEO Michael Truell responded quickly on X, noting that OpenAI models currently serve only about 5 percent of Cursor's actual user traffic and that the two companies are actively discussing how to resolve the situation, while calling Cursor one of OpenAI's very first users and noting they'd trusted OpenAI's platform to remain neutral infrastructure for their business for years. Anthropic, notably, moved quickly to position itself as the more reliable partner in the wake of the news, with co-founder Tom Brown publicly writing that Cursor has been a trusted partner since Claude Sonnet 3.5 and that Anthropic will keep expanding compute capacity for Claude models inside Cursor specifically

Cursor has also been building its own competing coding model, Composer 2.5, based on an open source checkpoint, which reduces its overall reliance on any single external frontier model provider. Curious what people think this specific episode reveals about how AI labs are treating model access as leverage in broader corporate and personal disputes

Our Decision On Cursor Following Its Acquisition By Spacexopenai.com

[Sabu_ECW]

Musk and Altman's ongoing feud clearly extends well beyond their own two companies directly at this point, now actively affecting genuinely unrelated third party businesses like Cursor that just happen to be caught in the middle of it.

That's a real concerning precedent for how personal corporate rivalries can end up spilling over unpredictably onto other companies entirely

[RatedRMike93]

Anthropic moving quickly to publicly position itself as the more reliable and trustworthy partner right in the immediate wake of this is honestly smart, fast competitive positioning regardless of the underlying dispute's actual merits between OpenAI and Musk specifically

[DarkMatter92]

Only five percent of traffic genuinely being at risk makes this feel more like a symbolic and personal statement against Musk specifically than a decision with real meaningful business consequences for Cursor's actual product or users. The number itself does most of the work undercutting how serious this decision initially sounds

[Seb5]

Citing specific past contract violations by name as the actual stated justification here, rather than something vaguer like strategic differences, is a genuinely notable and unusually direct move for a major company to make publicly. Feels like OpenAI wanted this reasoning documented clearly and on the record rather than staying deliberately vague about it

[RightNutter]

Building Composer 2.5 as a genuine hedge against exactly this kind of single provider dependency risk now looks like a remarkably prescient decision on Cursor's part in hindsight. Diversifying your core model providers early clearly paid off here when the actual risk eventually materialized