OpenELM refers to Open-source Efficient Language Models
Apple had said that the OpenELM AI model was trained on public data
The AI models were designed for text-related tasks
Apple’s OpenELM AI models have up to 3 billion parameters
Photo Credit: Reuters
Apple has reportedly claimed that the Open-source Efficient Language Models (OpenELM) AI model it released recently was not used for its Apple Intelligence or any of the company's machine learning features. The Cupertino-based tech giant open-sourced the OpenELM model in April. The reported statement comes just days after an investigation found that the AI model was trained on datasets that contained subtitles from hundreds of thousands of YouTube videos. Notably, these datasets were also used to train AI models of Anthropic, Salesforce, Nvidia, and more.
Apple Reportedly Denies Using OpenELM for Apple Intelligence
Speaking with 9to5Mac, the iPhone maker said that OpenELM does not power either of its AI features under the Apple Intelligence branding or other machine learning features. The company also added that OpenELM was created as a contribution to the research community to help them advance the development of large language models, as per the report.
Since the AI model was designed for research purposes, it was not used for any of the in-house AI innovations of Apple, the company told the publication. Notably, OpenELM is still open-sourced, and the model weights can be accessed from the company's Hugging Face listing.
At the time of release, Apple made the OpenELM family of AI models available in two variants — pre-trained and instruction. These models are further separated into different parameters of 270 million, 450 million, 1.1 billion, and 3 billion.
AI Models Trained on YouTube Data
A recent investigation accused Apple's OpenELM and multiple other AI models by major AI firms of using subtitles data from YouTube videos for training. The report found that companies such as Anthropic, Salesforce, Nvidia, and others used a publicly available data repository called the Pile, which contained data from YouTube videos of major content creators such as Marques Brownlee, MrBeast, CarryMinati, and more.
While Apple denied the usage of the AI model in its devices and operating systems, Anthropic's spokesperson Jennifer Martinez told Proof News, the publication that conducted the investigation, “The Pile includes a very small subset of YouTube subtitles. YouTube's terms cover direct use of its platform, which is distinct from use of the Pile dataset. On the point about potential violations of YouTube's terms of service, we'd have to refer you to the Pile authors.”
Apple Said to Have Airlifted 600 Tons of iPhones From India to 'Beat' Trump Tariffs11 April 2025
Search
![Gadgets 360 With Technical Guruji: News of the Week [April 12, 2024]](https://c.ndtvimg.com/2025-04/jlpk5kco_news-of-the-week_160x120_12_April_25.jpg?downsize=180:*)
![Gadgets 360 With Technical Guruji: Ask TG [April 12, 2024]](https://c.ndtvimg.com/2025-04/150s7qio_ask-tg_160x120_12_April_25.jpg?downsize=180:*)
