Japan to Require AI Firms to Disclose Training Data
Japan is preparing a nonbinding "comply or explain" code that would urge generative AI companies, including foreign firms operating in Japan, to disclose what models they use, what training data they rely on, and how that data was collected. The proposal would also let rights holders ask whether specific webpages were included in training datasets. The Japan Times reports: The draft code…
Japan is drafting a nonbinding "comply or explain" code that would compel generative AI companies, including those operating abroad in Japan, to disclose vital information regarding their AI models, training data, and collection methods. The proposal extends to rights holders who can request whether particular webpages were incorporated into the training datasets.
According to The Japan Times, the draft code is structured around three core principles. The first principle mandates companies to openly reveal the generative AI models they utilize, the data they train on, and the methods employed for data acquisition on their websites, without mandating the disclosure of sensitive information.
The second principle requires businesses to disclose when specific webpages are part of the training data upon request from copyright holders and rights holders alleging infringement. The third principle dictates that companies must respond to inquiries from users of their systems and services who are concerned about copyright infringement.
Written by urgent.news from Slashdot's reporting — not their text. Machine-written — may contain errors; check the original before relying on it.
This story
This is one outlet's version. Read the fullest account.
