Apple accused of scraping millions of youtube videos for ai training

A new lawsuit alleges Apple bypassed YouTube's protections to download millions of videos and train its AI model. Three YouTube creators claim Apple used their content without permission, violating privacy and copyright.

A proposed class action against apple

The lawsuit, filed by Ted Entertainment, Matt Fisher, and Golfholics, accuses Apple of scraping over 500 of their videos to create a massive index called Panda-70M. This dataset was used to train Apple's video generation AI model, as described in a research paper published last year.

Apple's AI efforts have faced scrutiny in recent times, with the company playing catch-up in the field. The irony lies in Apple's reputation for prioritizing privacy, which makes this alleged data scrape even more egregious.

Amazon and OpenAI also face similar lawsuits over the same dataset, highlighting an industry-wide issue: tech giants treating creator content as free fuel for AI development without proper licensing or consent.

Creators are pushing back against these practices, with publishers already limiting Apple's web crawlers and individual creators now joining the fight. The question remains: where do you draw the line when it comes to using online content for AI training?