YouTube, the video-sharing platform, has urged OpenAI to refrain from using videos published on the platform in the development and training of artificial intelligence systems it creates.
Neil Mohan, the CEO of YouTube, stated in a interview with the American newspaper Bloomberg that using site videos to train the AI models developed by OpenAI would be a clear violation of the terms of use that have been set.
Mohan said: “From a content creator’s perspective, when they put in effort and publish on our platform, certain things are expected, including adhering to the terms of service that prohibit uploading videos or accompanying texts illegally, which would be a clear breach of our terms of service. These are the laws that govern content on our platform.”
OpenAI, specializing in artificial intelligence, recently announced a new model called “Sora,” which converts written instructions into short videos using AI technologies.
The unveiling of the Sora model sparked various conflicting positions due to the ambiguity surrounding it from OpenAI and its failure to clarify the sources of data used in developing the model.
In March of last year, the company’s Chief Technology Officer, Mira Morati, played a role in exacerbating the criticisms facing the company when she stated in an interview with The Wall Street Journal her uncertainty about whether the Sora system adopted by the company uses data from sites like YouTube, Instagram, or Facebook.
Recently, The Wall Street Journal reported that OpenAI intends to leverage the texts accompanying videos on YouTube in the development and training of its next version of the artificial intelligence model, GPT-5.
On the other hand, Google, a major competitor of OpenAI in the field of artificial intelligence, claims to adhere to contractual obligations related to the use of the YouTube platform, owned by the company. Google’s artificial intelligence model, Gemini, requires similar data for training and development purposes.
Mohan claimed that Google’s use is limited to videos that have obtained proper permissions according to the licensing contracts of each content creator.
