Musk at first Cursor all-hands concedes Anthropic leads the AI race
At his first all-hands as Cursor's new owner, Elon Musk told employees exactly where he thinks Grok stands — and where the frontier sits without him. The session, reported by The Information, carried a strikingly rare note of public humility from a founder who usually beats the drum: Musk said Grok still needs to catch up, that Anthropic is leading the AI race, and that AI is heading toward a point where it will become impossible for humans to control.
Musk used his first Cursor all-hands to concede Grok is chasing both Anthropic and the day AI slips beyond human control. Speaking to the coding-tool team SpaceX absorbed in a record $60 billion all-stock deal two weeks ago, the billionaire reportedly told staff that Grok "needs to catch up," named Anthropic "the AI race" leader, and warned that AI will eventually become impossible for humans to control. The framing is notable on two fronts: it is a surprisingly direct admission from the owner of the model family that Cursor is being pivoted toward — and it doubles as a safety warning that clashes with the aggressive, go-fast posture Musk has generally taken on AI timeline questions.
The substance lands against a specific backdrop. SpaceX closed its acquisition of Cursor (Anysphere) on August 14, folding it into the rebranded SpaceXAI division, with the editor expected to default toward Grok models rather than the Claude and GPT models developers have relied on there. So a public "we're behind" from the man now steering Cursor is both a morale message to the team absorbing that pivot and a strategic signal about what to expect next: Musk cannot plausibly claim his own model is a safe hand on an uncontrollable frontier while simultaneously conceding his product is the laggard in the race to build it. The tension between those two statements — we're behind, and nobody will be able to control it anyway — is the real story.
We covered the close of that deal last week — SpaceX completes record $60B Cursor acquisition. The trust test is exactly what Musk's comments stress: he's asking the developers who trusted Cursor as a neutral editor to underwrite a Grok-led roadmap at the same moment he publicly frames that technology as an uncontrollable risk. It is a hard sell — and one Anthropic, holding the leader position Musk just conceded, is positioned to exploit.
What to watch: whether Cursor's Claude and GPT model options survive the pivot, and how fast Grok climbs from "catch up" to default.
If even the owner of a $60 billion coding tool admits AI will outgrow human control, should developers trust that tool with their work — or hedge now? Tell us in the comments.
