OpenAI had announced plans for its ‘Media Manager’ tool seven months ago, aiming to allow creators and content owners to opt-out of having their work used in AI training. The goal was to provide a standard for identifying copyrighted text, images, audio, and video across multiple sources.
However, insiders claim that the company has downplayed the feature as a priority since its initial announcement in May 2024. TechCrunch spoke with former OpenAI employees and someone coordinating work with the company, revealing that there have been no recent updates on Media Manager.
This comes at a time when OpenAI faces criticism over its approach to using copyrighted content in training its models. The Sora model, for instance, can generate material featuring copyrighted characters, works, and compositions due to its vast training data.
The AI’s inability to distinguish between original and altered content has made it challenging for users to request exact matches of copyrighted content. OpenAI is currently dealing with class action lawsuits from artists, writers, news organizations, and YouTubers who claim the company used their work without permission.
Despite the initial promise, Media Manager remains unavailable, sparking concerns over OpenAI’s commitment to respecting creators’ rights.
Source: https://www.digitalmusicnews.com/2025/01/01/openai-fails-to-deliver-opt-out-tool-it-promised-in-2024