news

한어Русский языкEnglishFrançaisIndonesianSanskrit日本語DeutschPortuguêsΕλληνικάespañolItalianoSuomalainenLatina

it home reported on august 28 that today, the california legislature passed a controversial bill - the artificial intelligence training data transparency act. the bill will be submitted to governor gavin newsom, but he has not yet made a statement on the bill.

image source: pexels

the bill aims to require ai companies to make more transparent disclosures about the data sets they use when developing and training models. it home noted that the bill (ab 2013) proposed by assemblyman jacques irving originally had a wider scope of application, but on august 20, the scope was narrowed to apply only to generative ai, that is, ai that can generate text, images, and similar content, such as open ai's chatgpt.

the bill requires ai companies to disclose key information about their training models, such as data sources, data types, and data processing methods, which is of great significance for protecting data privacy and preventing data abuse, and also helps the public better understand ai technology. however, the industry is concerned that overly strict transparency requirements may have an adverse impact on corporate trade secrets and technological innovation.