Elon Musk Admits xAI Has Used OpenAI Models for Training

In a startling admission during the ongoing Musk v. Altman trial, Elon Musk testified under oath that his AI company, xAI, has utilized OpenAI’s models to train its own systems. The revelation came as Musk was questioned about xAI’s development practices, and he defended the move by arguing that such cross-usage is standard procedure across the AI industry. “Every major lab learns from the work of others,” Musk stated during his testimony, adding that model training often involves benchmarking against publicly available outputs. However, the admission carries significant legal weight given that Musk is suing OpenAI for allegedly abandoning its nonprofit mission and engaging in anticompetitive behavior. Critics were quick to point out the apparent contradiction: Musk is accusing OpenAI of unfair practices while simultaneously benefiting from its technology. The disclosure adds a complex new dimension to the trial, highlighting the tangled web of interdependencies that define the modern AI landscape. Many AI companies, including xAI, rely on open-source models and publicly accessible APIs to accelerate their research. Yet, this practice raises thorny questions about intellectual property, attribution, and fair competition. Legal analysts suggest that Musk’s admission could weaken his case, as it demonstrates that xAI itself is not operating in a vacuum. On the other hand, it also underscores the industry-wide challenge of establishing clear boundaries for model usage. As the trial continues, this revelation is likely to fuel further debate about whether current AI development norms are sustainable—or if new rules are needed to govern how companies build upon each other’s work.