In May 2024, OpenAI announced plans to introduce a tool called "Media Manager" that would allow creators to control how their work is used in AI training data. The promise was hailed as a step toward addressing mounting concerns over intellectual property (IP) rights in the age of generative AI. However, seven months later, the tool remains unavailable, raising questions about the company's commitment to creators and compliance with its self-imposed deadline to launch the feature “by 2025.”

OpenAI’s Media Manager was envisioned as a robust platform capable of identifying and managing copyrighted text, images, audio, and video. It aimed to reflect creators’ preferences across multiple content types, potentially shielding OpenAI from legal challenges while placating critics. At the time, the company described it as an effort to “set a standard across the AI industry.”

However, sources cited by TechCrunch suggest that the tool was not prioritized internally. One former OpenAI employee commented, “I don’t think it was a priority... I don’t remember anyone working on it.” Additionally, Fred von Lohmann, a member of OpenAI’s legal team involved in Media Manager’s development, transitioned to a part-time consultant role in October. OpenAI confirmed the move to TechCrunch but has remained silent on the tool’s progress.

A Legal and Ethical Quagmire

The absence of Media Manager comes as OpenAI faces a growing number of lawsuits alleging that its AI models were trained on copyrighted materials without permission. Plaintiffs, including authors, visual artists, and news organizations, argue that the company’s models regurgitate near-copies of their work, blurring the line between fair use and infringement.

While OpenAI has introduced limited measures, such as a submission form allowing creators to opt out of future training sets, critics have labeled these efforts as insufficient. According to TechCrunch, the opt-out process for images, for example, is cumbersome and lacks mechanisms for other media formats like video and audio.

Ed Newton-Rex, founder of Fairly Trained, told TechCrunch that the burden of using tools like Media Manager unfairly shifts responsibility onto creators. “Most creators will never even hear about it, let alone use it,” he said.

Even if Media Manager were to launch, experts doubt its ability to resolve the deeper legal questions surrounding AI training and IP usage. Speaking to TechCrunch, Adrian Cyhan, an IP attorney, noted the significant challenges in developing a system capable of managing content at scale, particularly given the diverse and rapidly evolving legal frameworks globally.

Further complicating matters, copyright law does not require creators to proactively prevent infringement, as noted by Evan Everist, a copyright law expert. Any tool OpenAI develops may serve more as a public relations gesture than a meaningful legal safeguard.