YouTube CEO Reminds OpenAI
YouTube
– YouTube CEO Neal Mohan warned OpenAI not to use video content from its platform to train AI models because this clearly violates the platform's terms of use.
According to an Engadget broadcast on Friday (5/4), the statement was made in response to OpenAI CTO Mira Murati, who stated at the end of March 2024 that she was not sure whether Sora's AI model took data from uploads on platforms such as YouTube, Instagram and Facebook.
In an interview with Bloomberg Originals, Mohan said that one of the expectations creators have when uploading work to YouTube is that its terms of service will be adhered to.
“It doesn't allow things like transcripts or video bytes to be downloaded, and that's a clear violation of our terms of service. Those are the rules when it comes to content on our platform,” Mohan said.
The problem of using individuals' work without permission (or compensation) by AI models is not new. Entities such as The New York Times and Getty Images have filed lawsuits against artificial intelligence (AI) creators along with artists and writers.
Uncertainty and controversy still surround how OpenAI trains Sora, along with ChatGPT and DALL-E.
The Wall Street Journal recently reported that the company founded in 2015 plans to use YouTube video transcriptions to train GPT-5.
As a competitor to OpenAI, Google seems to follow the rules more, at least when it comes to using content on YouTube, its platform.
Google's Gemini AI model, according to Mohan, only uses certain videos, depending on permissions granted under licensing contracts with individual creators.
source: Antara