REGULATORY · COMPETITIVE · USA
Court denies xAI bid to block AB 2013
Change
A court denied xAI’s attempt to block enforcement of California’s Assembly Bill 2013 requiring AI developers with models accessible in the state to disclose specified training-data sourcing details.
Why it matters
The law applies to AI developers whose models are accessible in California. Required disclosures include which dataset sources were used to train models, when the data was collected, and whether collection is ongoing. Disclosures must state whether datasets include data protected by copyrights, trademarks, or patents, and whether training data was licensed or purchased. Disclosures must also address whether training data included any personal information and how much synthetic data was used to train the model.
Implications
- • AI developers must compile and publish training-data sourcing and provenance details for models accessible in California.
- • Developers must disclose whether training data includes personal information and whether it includes copyrighted, trademarked, or patented material.
- • Developers must disclose whether training data was licensed or purchased and the extent of synthetic data used in training.
Who is affected
- • AI model developers with products accessible in California
- • Legal and compliance teams responsible for statutory disclosures
- • Data sourcing and procurement teams managing training-data licensing and purchase
- • Data governance teams tracking dataset provenance and collection timing
Source
Topics
World & Politics Policy & Regulation Law & Public Safety Data Privacy Technology & Innovation Artificial Intelligence