Apple , NVIDIA , and Anthropic have allegedly used YouTube videos without permission to train their AI models.
A recent investigation by Proof News reveals that some of the world's largest tech companies trained their artificial intelligence models using a dataset that included conversions from more than 173,000 YouTube videos, without permission.
The dataset, created by the nonprofit EleutherAI, includes video conversions from more than 48,000 YouTube channels and has been used by companies like Apple, NVIDIA, and Anthropic. The research results reveal an uncomfortable truth about artificial intelligence: the technology relies heavily on data collected without the consent or compensation of its creators.
Read also: Apple: What to expect from the iOS 18 public beta?
The dataset does not include videos or images from YouTube, but it does contain video conversions from top YouTube creators like Marques Brownlee and MrBeast, as well as major news publishers like the New York Times, BBC, and ABC News. Subtitles from Engadget videos are also included.
“Apple has collected data for its artificial intelligence from many companies,” Brownlee told X. “.
A Google spokesperson told Engadget that YouTube CEO Neil Mohan's previous comments that companies using YouTube data to train AI models violate the platform's terms and services remain in place. Apple, NVIDIA, Anthropic, and EleutherAI did not respond to a request for comment from Engadget.
So far, AI companies have been relatively transparent about the data they use to train their models. Earlier this month, artists and photographers criticized Apple for failing to disclose the source of the training data for Apple Intelligence, its own version of Generative AI that will be built into millions of Apple devices this year.
See more: How to download videos from YouTube on Mac
YouTube, the world’s largest video repository, is a goldmine not only for conversions, but also for audio, video, and images. This makes it an invaluable source of data for training artificial intelligence (AI). OpenAI’s chief technology officer, Mira Murati, declined to answer questions from the Wall Street Journal about whether the company uses YouTube videos to train Sora ,OpenAI’s upcoming AI video creation tool.
“I won’t go into the details of the data used, but it was either publicly available or licensed,” Mira Murati said at the time.
Alphabet CEO Sundar Pichai also said that companies using data from YouTube to train their artificial intelligence models are violating the platform's terms of service.
Read also: YouTube Music and Amazon Music are coming to Tesla EVs
To check if subtitles from your YouTube videos or favorite channels are included in the database, visit the Proof News search tool.
Source: engadget
