Elon Musk Confirms xAI Used OpenAI Models to Train Grok
Elon Musk testified in federal court that his AI startup, xAI, utilized OpenAI's models to improve its own Grok chatbot through a process known as model distillation
Elon Musk testified in a California federal courtroom on Thursday that his artificial intelligence startup, xAI, utilized models from OpenAI to improve its own technology. This admission confirms that Grok, xAI's flagship AI model, was developed in part by leveraging the outputs of a major competitor. Musk's testimony provides a rare look into the training methods used by xAI as it seeks to compete with more established players in the generative AI space.[1][2]
The process described by Musk is known as model distillation, a technique where a larger, more advanced teacher model is used to train a smaller student model. While distillation is a common industry practice for companies looking to optimize their own internal models, it is also frequently employed by smaller labs seeking to replicate the performance of leading frontier models. By using OpenAI's models as a reference, xAI was able to pass on knowledge to its own smaller systems.[1][2]
The use of distillation has become a significant point of contention within the AI industry as frontier labs try to prevent smaller competitors from copying their models. While the practice can be used legitimately within a single company to refine its own tools, it is also a method for startups to mimic the performance of larger competitors. Musk's confirmation of this practice at xAI underscores the competitive pressures and technical shortcuts prevalent in the race to develop advanced artificial intelligence.[1][2]
