FTFuture Technology
AI

OpenAI Is Cutting Off Cursor After SpaceX Bought It

· 3 min read · By Nath Connell

Key takeaways

  • OpenAI is terminating its model supply contract with Cursor following SpaceX's acquisition of the coding tool
  • Cursor previously integrated OpenAI, Anthropic Claude, and other model providers
  • Elon Musk founded xAI, a direct OpenAI competitor, and has been in public legal conflict with OpenAI since 2018
  • The move signals that API access to frontier models can be revoked based on corporate ownership, not just contract terms

This one has a few layers to unpick. OpenAI has announced it is winding down its contract to provide AI models to Cursor, the popular AI coding assistant, following Cursor's acquisition by SpaceX. The decision lands at a genuinely awkward intersection of commercial competition, national security concerns, and the increasingly tangled web of Elon Musk's business relationships.

Cursor built its reputation as one of the most capable AI-powered code editors on the market, integrating models from multiple providers to give developers a seriously impressive autocomplete and code generation experience. Its acquisition by SpaceX, Musk's private space and technology company, changes the calculus for OpenAI significantly.

Why OpenAI Is Walking Away

OpenAI's statement was careful and measured in the way that corporate communications about awkward situations usually are, but the core logic is not hard to follow. OpenAI and Elon Musk have been in open conflict since Musk departed OpenAI's board in 2018 and subsequently launched his own AI company, xAI. Musk has been a vocal and litigious critic of OpenAI, and the relationship between his various companies and OpenAI's technology has been a source of ongoing tension.

Providing models to a Musk-controlled company would put OpenAI in the strange position of powering a competitor's tooling. SpaceX is not just any acquirer. It has contracts with the US Department of Defense, runs Starlink, and sits at the centre of a commercial space industry that overlaps with national security infrastructure. The concern, at least on the surface, is about who ultimately controls the systems that OpenAI's models are embedded in.

There is also the less politely stated version: OpenAI simply does not want to help a Musk entity thrive, and it now has a legitimate business reason not to.

What Happens to Cursor Now

Cursor's users are understandably anxious. The editor has a substantial and dedicated developer base, many of whom chose it specifically for the quality of the model integrations. Losing OpenAI's models, which include GPT-4 class systems that many users treat as the baseline for serious coding tasks, is not a trivial blow.

The future, in 3 minutes a day. The biggest tech story explained every morning, free. Get the briefing →

The good news, such as it is, is that Cursor has always been multi-model. It also integrates Anthropic's Claude and can route to other providers depending on the task. SpaceX has significant resources and Musk has his own AI company in xAI, whose Grok models could theoretically fill the gap. Whether Grok-based completions will satisfy developers who are used to GPT-4 quality is a different question, and one that the developer community is already debating loudly.

There's also a scenario where SpaceX migrates Cursor toward xAI's infrastructure entirely, which would make the product a showcase for Musk's own models and reduce any dependency on third-party AI providers. From a strategic standpoint, that's probably the direction this goes.

The Bigger Picture

What this story really illustrates is how the AI tooling market is becoming a battleground shaped by corporate allegiances as much as technical capability. Developers building serious products are increasingly forced to think about not just which AI model is best for their task, but whose model it is, who controls the company that makes it, and what commercial or political relationships might disrupt the supply chain.

OpenAI's decision sets a precedent, too. It's effectively using API access as a competitive and geopolitical lever. That might be entirely reasonable in this specific case, but it signals to the broader market that access to frontier AI models is conditional, not just on payment, but on who you are and who owns you.

For developers, the lesson is probably one the open-source community has been making for years: dependency on proprietary model providers is a risk. The more your product is built on API calls to a company that can pull the plug for reasons entirely unrelated to your performance as a customer, the more vulnerable your product is. That's not a comfortable position to be in, and this situation won't be the last time it causes problems.

Get the briefing, free

The biggest tech story, explained in 3 minutes every weekday. Choose your briefings →

Free. No spam. Unsubscribe in one click.