Sam Altman and Elon Musk seem to always be at odds. Just when it seemed that both tech moguls had calmed down after their bitter squabble over social media, Altman’s OpenAI has taken a decision that could reignite the matter.
OpenAI officially notified SpaceX that it will terminate its long-standing model-supply agreement with Cursor, the market-leading AI coding platform developed by Anysphere and now owned by SpaceXAI. The trigger? SpaceX’s recent $60 billion acquisition of Anysphere—and, more specifically, OpenAI’s explicit refusal to do business with companies controlled by Elon Musk.
The OpenAI-Cursor breakup
OpenAI’s decision to cease supply of its models brings an abrupt end to a four-year commercial partnership with Cursor. By invoking a standard “change of control” clause in its agreement, OpenAI has set a hard deadline, stating that direct API access to its AI models will terminate for Cursor on November 12, 2026.
Unlike standard corporate realignments, which are typically framed around product shifts or non-competes, OpenAI did not shy away from expressing its actual intentions. In a public blog post detailing the decision, the Altman’s AI lab explicitly named Elon Musk as the primary cause.
“We are making this choice because we cannot be confident that SpaceX will use our technology within our terms of service, based on our experience with Elon Musk’s companies violating contracts,” OpenAI wrote.
The post referred to prior contractual breaches by X (formerly Twitter) following Musk’s 2022 takeover, while also citing sworn testimony from earlier this year in which Musk acknowledged that xAI, which has now been consolidated under the SpaceX umbrella, had used OpenAI data to train its own models in violation of service terms.
“After Musk acquired Twitter, now part of SpaceX, the company broke(opens in a new window) the terms of our contract (alongside many others). Under oath earlier this year, Musk admitted(opens in a new window) that xAI, now also part of SpaceX, had violated OpenAI’s terms of service (terms which are similar to xAI’s own),” wrote the company.
Furthermore, OpenAI signalled that its upcoming flagship model, codenamed Astra, comes with strict safety and deployment protocols that governance leadership was unwilling to extend to a Musk-controlled entity.
How Musk fired back
The fallout on social media was swift, as expected, highlighting both the personal nature of the conflict and the high stakes for the software engineering market.
Elon Musk fired back on X, dismissing the loss of access. “I couldn’t care less. Scam Altman and Greg Stockman are utterly untrustworthy a******* who stole an open source nonprofit.”
Michael Truell, the co-founder of Cursor and now a SpaceX executive, stuck to a more measured tone. He noted that OpenAI models currently handle roughly 5 per cent of Cursor’s traffic, as the editor increasingly relies on its proprietary models and native compute backed by SpaceX’s vast GPU infrastructure. “OpenAI models serve about 5% of Cursor user traffic, and we’re speaking with the OpenAI team to resolve this, wrote Truell, adding that, “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.”
Anthropic wasted no time stepping into the void. Tom Brown, Anthropic’s Chief Compute Officer, announced that the company would immediately expand compute support for its Claude model suite within Cursor, a post Musk endorsed on X with two rocket emojis.
OpenAI’s controversy amid Musk-Altman feud
OpenAI’s decision comes on the heels of the recent high-stakes tension between Elon Musk and Sam Altman.
Musk sued OpenAI, Sam Altman, and President Greg Brockman for$150 billion, accusing them of “stealing a charity” and unjustly enriching themselves by transforming a non-profit research lab into a commercial powerhouse. Musk’s legal team argued that his early $44 million in donations were applicable on the condition of OpenAI remaining dedicated to open-source safety. Musk’s team painted Altman as a duplicitous actor who engineered a corporate bait-and-switch.
OpenAI, meanwhile, fired back with internal documents showing that Musk himself had championed a for-profit pivot as early as 2017, demanding majority equity and board control for himself before walking away when the board refused.
In May 2026, after weeks of theatrical courtroom testimony, the nine-person federal jury reached a verdict in favour of Altman and OpenAI. The jury concluded that Musk had delayed unreasonably in bringing his claims, ruling that the lawsuit fell outside the legal statute of limitations.
How Cursor users are affected
API key workarounds: Individual developers using Cursor can still bring their own OpenAI API keys or use OpenAI’s official IDE extensions, meaning GPT models won’t disappear completely from developer screens.
Shift to Grok and Claude: With Cursor integrating heavily into SpaceX’s compute infrastructure (including its Colossus GPU cluster) and running native models like Grok 4.6 alongside Anthropic’s Claude, the crow relying on OpenAI models can shift to the readily available models.
