A YouTube creator has filed a class action lawsuit against OpenAI over the unauthorized use of his videos in education is located.
In a report published last week, it was reported that many YouTube videos were used without permission for training the Apple Intelligence infrastructure. In the prepared report, It was stated that more than 170 thousand videos were used in the training of large language models underlying the Apple Intelligence infrastructure introduced at WWDC24. In the process, subtitles in videos of names such as Marques Brownlee (MKBHD), MrBeast, PewDiePie, Stephen Colbert, John Oliver and Jimmy Kimmel, It was acquired by a non-profit organization called EleutherAI. As will come as no surprise, permission was not obtained from the content creators. Marques Brownlee, who made a statement on the controversial issue where companies such as Nvidia and Anthropic were mentioned, said, “Apple has sourced data for its AI systems from various companies. One of these companies has collected tons of data/transcripts from YouTube videos, including mine. Apple technically avoids this situation/issue because it is not collecting the subtitles. However, this usage will be a serious problem in the long run.” he had said. The person who brought this issue to the agenda now is YouTuber David Millette It happened. Millette, who filed a class-action lawsuit against OpenAI, alleged that OpenAI secretly used videos of himself and other creators to train the models that power its AI-powered chatbot ChatGPT. Millette, who is represented by the law firm Bursor and Fisher, said: It is seeking over $5 million in damages on behalf of all YouTube users whose data OpenAI may have used in training.
YOU MAY BE INTERESTED IN
An official statement on this matter later making a statement Applestated that none of the models trained on YouTube videos power the Apple Intelligence infrastructure. In the shared news, YouTube videos OpenELM Apple, which is said to have used a model called OpenELM to train, stated in its statement that OpenELM was created for research purposes only and that the model is open source to everyone. The technology giant, which also stated that it will not develop a new version of OpenELM, did not say in its statement that YouTube videos were not used in the training process. In other words, this “unauthorized” situation was clearly accepted.
Runway AI was also on the agenda last week on this issue. According to 404media’s research According to Runway AI’s OpenAI Sora rival artificial intelligence system, which can generate videos from text, was trained using popular videos from YouTube channels, as well as content from companies such as The New Yorker, VICE News, Pixar, Disney, Netflix and Sony. The company that allegedly used these contents without permissionremained at the center of great criticism on this issue.