Earlier this week, an investigation detailed that Apple and other tech giants had used YouTube subtitles to train their artificial intelligence models. This includes more than 170,000 videos from MKBHD, MrBeast, and more. Apple then used this dataset to train its open-source OpenELM models, which it released back in April.
However, Apple has now confirmed that OpenELM does not support any of its artificial intelligence or machine learning features, including Apple Intelligence.
Apple said it created the OpenELM model to contribute to the research community and advance the development of open-source large-scale language models. In the past, Apple researchers have described OpenELM as the "state-of-the-art open language model." According to Apple, OpenELM is for research purposes only and is not used to support any of its Apple Intelligence features. The model is released as open source and is widely available, including on Apple's machine learning research website.
Since OpenELM is not used as part of Apple Intelligence, this means that the "YouTube Subtitles" dataset was not used for Apple Intelligence training. In the past, Apple has said that Apple Intelligence models are trained "on licensed data, including data selected to enhance specific features, as well as publicly available data collected by our web crawlers."
Finally, Apple also stated that it has no plans to build any new versions of the OpenELM model.
As Wired reported earlier this week, companies including Apple, Anthropic, and Nvidia have used this "YouTube captions" dataset to train their artificial intelligence models. This dataset is part of a larger dataset called "ThePile" from the non-profit organization EleutherAI.