California has moved a step closer to implementing one of the most detailed legislative frameworks for the transparency of AI training data. On Tuesday, the California legislature passed the AI Training Data Transparency Bill, officially known as AB 2013. Authored by Assemblymember Jacqui Irwin (D), the bill mandates that companies disclose summaries of the data sets used to train their generative AI models. This would encompass AI tools capable of creating content such as text and images, exemplified by OpenAI’s ChatGPT.
The bill now awaits the decision of Governor Gavin Newsom, who has yet to express his stance on the measure. If signed into law, AB 2013 would position California at the forefront of AI regulatory standards in the United States, surpassing the existing provisions found in Colorado’s AI law.
Supporters argue that the law is a necessary step toward accountability and consumer protection. However, the tech industry has voiced concerns about potential trade secret exposures and competitive disadvantages. These reservations were a focal point of debate leading to the bill’s passage.
The bill is seen as an ambitious effort to strike a balance between fostering innovation and ensuring transparency. According to analysis, the ultimate implementation and enforcement of this bill could significantly impact AI operations not only within California but also set a precedent for other states and possibly at the federal level.
For additional details, you can read the full article on Bloomberg Law.