March 9, 2026 A U.S. federal judge has denied Elon Musk’s xAI request for a preliminary injunction that would have temporarily blocked California from enforcing a law requiring AI developers to disclose information about the data used to train their models. The ruling means companies offering AI systems in the state must comply with the transparency requirements while the lawsuit continues.
The law, known as Assembly Bill 2013 (AB 2013), requires AI developers to publicly describe key details about their training datasets, including the sources of the data, when it was collected, whether collection is ongoing, and whether the datasets contain copyrighted material or personal information. The disclosures must also indicate whether training data was licensed or purchased and whether synthetic data was used.
xAI argued the law forces companies to reveal sensitive trade secrets. In its complaint, the company said disclosing dataset sources, dataset sizes and data-cleaning practices could expose competitive advantages and potentially allow rivals to replicate its training strategy.
U.S. District Judge Jesus Bernal rejected the request, saying xAI failed to demonstrate that the law requires disclosure of protected trade secrets. In his order, Bernal wrote that the company relied largely on broad claims about the importance of training data without identifying specific harms that would result from the disclosures.
The judge also rejected xAI’s constitutional arguments, which claimed the law violated the Fifth Amendment by forcing disclosure of trade secrets and infringed on First Amendment protections by compelling speech from AI developers.
Bernal acknowledged that training data could hypothetically qualify as a trade secret but said xAI had not shown that its datasets or data-processing methods were uniquely sensitive compared with those used by competitors.
The court also sided with the state’s argument that the disclosure requirements serve a public interest by helping consumers evaluate how AI systems are trained. Bernal wrote that the law provides information that could help users determine whether they trust or want to rely on a given AI model.
The decision is an early procedural ruling, meaning the broader lawsuit will continue. However, the judge’s reasoning suggests that xAI faces a significant legal challenge if it seeks to overturn the law.
