Elon Musk’s xAI Challenges California’s AI Disclosure Law in Court

Elon Musk’s artificial intelligence company, xAI, has initiated a lawsuit against California’s Assembly Bill 2013, which requires generative AI developers to disclose detailed information about their training datasets. Enacted on September 28, 2024, this legislation, known as the Generative Artificial Intelligence: Training Data Transparency Act, is set to take effect on January 1, 2026. The lawsuit was filed in the Central District of California towards the end of 2025.

The complaint seeks to prevent California Attorney General Rob Bonta from enforcing the provisions of AB 2013 against xAI. The company argues that complying with the law would undermine its competitive edge by disclosing valuable trade secrets related to the datasets used in training its flagship AI chatbot, Grok. xAI asserts that it has invested significant resources in curating high-quality data, and that revealing this information would provide competitors with a blueprint for developing and training similar AI models.

xAI’s legal arguments center on the assertion that the disclosure requirements of AB 2013 violate the U.S. Constitution’s Takings Clause. The company claims that the law effectively forces it to relinquish its trade secrets or diminishes their value without any compensation. Additionally, xAI contends that the law infringes upon its First Amendment rights by compelling the company to disclose specific information, which it equates to a restriction on free speech.

The legislative intent behind AB 2013, which includes aims to “help identify and mitigate biases” in AI systems, is also a point of contention for xAI. The company argues that this focus on bias triggers a strict scrutiny standard, as it pertains to content- and viewpoint-based regulation of speech.

This lawsuit arrives at a time when xAI faces significant scrutiny, particularly concerning allegations that Grok generated sexually explicit content involving minors. A 2023 study from Stanford University identified thousands of child sexual abuse images within datasets used for training AI image generation models, raising ethical concerns about the technology’s implications.

With xAI valued at over $200 billion, the financial resources available for litigation on both sides are substantial. The implications of this case could set important precedents for the regulation of AI technologies and the balance between transparency and protecting intellectual property in a rapidly evolving field.

As the legal battle unfolds, the outcome may significantly impact not only xAI but also the broader landscape of generative AI development and regulation.