OpenAI, the prominent artificial intelligence research lab, has yet to deliver on its promise of developing a tool to let content creators control how their work is used in AI training data. The Media Manager tool, announced in May 2022, aimed to flag copyrighted materials but was never given priority. This has left critics and creators concerned about potential copyright infringement.
The company faces multiple class-action lawsuits from artists, writers, YouTubers, and media organizations alleging the use of their works without permission for AI training. Notable plaintiffs include authors Sarah Silverman and Ta Nehisi-Coates. OpenAI’s current opt-out methods have been deemed insufficient, with creators criticizing the lack of specific mechanisms for written works, videos, and audio recordings.
Experts are skeptical about Media Manager’s potential to fully address creators’ concerns or resolve the legal issues surrounding AI and IP usage. Adrian Cyhan, an IP attorney, notes that content identification at scale is a challenging task. Ed Newton-Rex, founder of Fairly Trained, worries the tool might unfairly shift the burden of controlling AI training onto creators. The delay in Media Manager’s development has sparked concerns about OpenAI’s commitment to protecting creators’ rights and addressing copyright infringement issues.
Source: https://www.newsbytesapp.com/news/science/openai-s-media-manager-tool-delayed-beyond-2025-deadline/story