SIAgentNews.com
← All news
DevelopingPolicy

Appeals court questions xAI's challenge to California's AI data disclosure law

Judges are scrutinizing Elon Musk's company's attempt to block a state law requiring AI developers to reveal their training data sources.

· First reported by MLQ.ai

12topics to be summarized in AI data disclosures

The short version

  • xAI is seeking to block California Assembly Bill 2013, which goes into effect on January 1, 2026.
  • The law requires developers of public generative AI to summarize 12 topics, including dataset sources and the use of copyrighted material.
  • A federal district judge previously denied xAI's request to block the law, leading to the current appeal.
  • Judges questioned if it is reasonable for the government to want citizens to know what information is being input into AI systems.

Why it matters

This case will help determine the balance between a company's right to protect trade secrets and the public's right to know how AI models are built. The outcome could set a precedent for how much transparency AI developers must provide to state governments.

The legal challenge in court

A federal appeals court recently heard a request from xAI, the artificial intelligence company owned by Elon Musk. The company is attempting to block a specific California law that mandates transparency regarding the data used to train AI models. As MLQ.ai reported, the company received a frosty response from the US Court of Appeals for the Ninth Circuit during the proceedings.

The court panel did not issue a final decision during this hearing but pressed both xAI and the state of California for further clarification. The judges are currently weighing the company's arguments against the state's regulatory goals. This legal battle marks a significant moment in the ongoing debate over how much information AI developers must share with the public.

xAI originally filed its lawsuit against California Attorney General Rob Bonta on December 29, 2025. The company argues that the state's requirements are unconstitutionally vague and violate the First Amendment by compelling speech. They also claim the law takes their trade secrets without providing any form of compensation to the company.

What the law requires

The specific legislation at the center of this dispute is Assembly Bill 2013. This law is scheduled to take effect on January 1, 2026, and it targets developers of publicly available generative AI systems. It requires these companies to provide a high-level summary of their training processes to ensure a level of public oversight.

Under the new rules, developers must cover at least 12 different topics in their summaries. These topics include the specific sources of the datasets used to build the models. The law also asks companies to disclose whether their training data contains personal information or copyrighted material.

Furthermore, the summary must clarify whether the data used by the company was purchased or licensed from other entities. By requiring these disclosures, California aims to create a clearer picture of the inputs that shape modern AI behavior. The state argues these requirements are necessary for public safety and transparency.

xAI's arguments and pushback

xAI's primary legal strategy rests on the idea that the disclosure requirements unreasonably regulate commercial speech. The company maintains that revealing the specifics of their training data could expose proprietary information. They argue that the government should not be able to force them to share these details as a condition of doing business in the state.

However, the judges expressed skepticism regarding this stance during the hearing. Judge Johnnie B. Rawlinson specifically questioned the company's position on government oversight. She asked if it is reasonable for the government to want citizens to know what is being input into an AI instrument, according to Bloomberg Law.

This line of questioning suggests that the court may be sympathetic to the state's interest in public knowledge. The judges are looking to see if the government's desire for transparency outweighs the company's desire for secrecy. The case highlights a fundamental tension between corporate privacy and public interest in the age of artificial intelligence.

The path to the current appeal

This legal battle did not begin with the appeals court. xAI first sought a preliminary injunction to stop the law from being enforced. They wanted a judge to step in immediately and block the requirements before the 2026 deadline arrived. The company argued they would suffer significant harm if the law moved forward as planned.

US District Judge Jesus Bernal reviewed the case and denied xAI's request for a preliminary injunction on March 4. The judge found that xAI had not demonstrated that it was likely to succeed on the merits of its legal arguments. This ruling was a significant setback for the company and prompted them to move the case to the appeals court.

The current proceedings in the Ninth Circuit represent the next level of judicial review. The appeals court will determine if the district judge's decision was correct or if the law should be struck down entirely. The outcome will determine whether xAI must comply with California's transparency rules or if they can keep their training data private.

Who is affected by the outcome

The resolution of this case will have significant implications for the entire AI industry. If the law stands, every developer of a public generative AI system in California will have to provide detailed summaries of their data. This could set a standard for other states to follow, creating a patchwork of different transparency requirements across the country.

For companies like xAI, the ruling will determine how much of their internal methodology they can keep hidden from competitors and the public. If they are forced to disclose sources, they may worry about the security of their intellectual property. They may also face increased scrutiny regarding how they handle copyrighted works and personal data.

For the general public, the ruling could lead to a greater understanding of how AI models are built. Knowing whether a model was trained on licensed data or scraped from the internet can affect user trust. The court's decision will ultimately shape the balance between technological innovation and the public's right to information.

What to watch

  • Whether the Ninth Circuit issues a ruling on xAI's First Amendment claims.
  • If other states adopt similar disclosure requirements for AI training data.

Sources and links

First reported by MLQ.ai. Written by the SI Agent News Desk from the reporting above.Written with AI from the source above. Every number is checked against the source. Spot a mistake? Corrections.

Liked it? Say why

Add a line for other readers.

No account needed. We never show your email. Comments with links, ads or abuse are held and not shown. Your comment shows after a quick check.

More AI news