Elon Musk confirms xAI used OpenAI’s models to train Grok

The confession came on April 30th, and it came under oath. Elon Musk testified that xAI used OpenAI's models during the training of Grok — a single sentence that lands with considerable force given everything surrounding it. Musk has spent years constructing himself as OpenAI's principal antagonist: the co-founder who departed, who sued the organisation he helped create, who accused it of abandoning its nonprofit mission while quietly building a commercial empire. That posture always had a theatrical quality to it, and now the curtain has shifted. The man who made the lawsuit also made a product using his opponent's technology. Both things are simultaneously true, which is the kind of contradiction that courtrooms are specifically designed to surface. The timeline matters here — the training of Grok predates xAI's public posturing as OpenAI's ideological alternative, but the timing of this admission, extracted in testimony rather than volunteered in a press release, suggests it would not have come out otherwise.

If confirmed — and at 87%, treat it as confirmed — this reshapes the entire framing of Musk's legal and rhetorical campaign against OpenAI. A lawsuit built on accusations of mission betrayal looks considerably different when the plaintiff was simultaneously using the defendant's models to build a competing product. For OpenAI's legal team, this is not just a point scored in a deposition; it is a potential argument about standing, about hypocrisy, about whether the grievance being litigated was ever entirely sincere. For the broader AI industry, it confirms something practitioners have long suspected but rarely said plainly: the frontier models are so expensive to train from scratch that even the companies loudest about independence are quietly standing on each other's shoulders. Grok was sold as a different kind of AI — less filtered, more honest, born outside the OpenAI ecosystem. That origin story has now been revised.

Watch for OpenAI's legal team to formally introduce this testimony into their defence strategy, and watch for any xAI clarification about which specific models were used and at what stage of Grok's development — the details there could either narrow or dramatically widen the implications.

NewsHive monitors these sources continuously. All signal titles above link to the original reporting.

Intelligence by NewsHive. Need help navigating what this means for your business? Contact GeekyBee →