OpenAI’s latest AI tool, Sora, which generates videos from text prompts, has sparked controversy over potential copyright infringement issues. The uncertainty surrounding how Sora is trained, including the use of YouTube content, has prompted a response from YouTube’s CEO Neal Mohan. Mohan made it clear that using YouTube videos to train Sora would violate the platform’s terms of service, which prohibit downloading video content or transcripts without permission.
In an interview with Bloomberg Originals, Mohan emphasized the importance of respecting creators’ work uploaded to YouTube and abiding by the platform’s terms of service. He stated that creators have certain expectations when sharing their content on YouTube, and unauthorized use of their materials is not allowed.
The situation highlights the ongoing debate surrounding AI models like Sora, ChatGPT, and DALL-E, as companies navigate the line between innovation and intellectual property rights. OpenAI’s plans to use YouTube video transcriptions to train its AI models have raised concerns about potential copyright violations. In contrast, Google, a competitor of OpenAI, has been noted for adhering to content usage rules, particularly with its AI model Gemini, which is designed to respect creators’ permissions.
HEALTH
WORLD AFFAIRS
TECHNOLOGY
CREATIVE
