AI firms are "illegally" scraping music, according to new investigation

International investigation reveals that major AI firms are allegedly training models on millions of copyrighted songs without license. This massive scale of data scraping suggests widespread, illegal appropriation of artistic work, challenging the assumption that such training is permissible. The findings highlight significant legal risks for tech companies, positioning current practices as severe copyright infringement. It forces a re-evaluation of how intellectual property laws apply to machine learning datasets, potentially leading to stricter regulatory oversight. This matters for open data because it underscores the critical need for transparent, licensed datasets. It illustrates the ethical and legal pitfalls of sourcing training materials from the open web, advocating for frameworks that respect creator rights in open data initiatives.

Source: mixmag.net
Published on 2025-09-04

AI firms are "illegally" scraping music, according to new investigation

Related news