OpenAI is fighting lawsuits from artists, writers and publishers who allege it inappropriately used their work to train the algorithms behind ChatGPT and other artificial intelligence systems. On Tuesday, the company announced a tool apparently designed to appease creatives and rights holders, giving them some control over how OpenAI uses their work.
The company says it will launch a tool in 2025 called Media Manager that will allow content creators to exclude their work from the company’s AI development. In a blog postOpenAI described the tool as a way to allow “content creators and owners to tell us what they own” and specify “how they want their work to be included or excluded from machine learning research and training.”
OpenAI said it is working with “creators, content owners and regulators” to develop the tool and intends to “set a standard in the industry.” The company did not name any of its partners in the project or make clear exactly how the tool will work.
Open questions about the system include whether content owners will be able to make a single request to cover all their jobs and whether OpenAI will allow requests related to models that have already been trained and released. Research is underway on machine “unlearning,” a process that tunes an AI system to retrospectively remove the contribution of a portion of its training data, but the technique has not yet been perfected.
Ed Newton-Rex, CEO of startup Fairly Trained, which certifies AI companies using ethically sourced training data, says OpenAI’s apparent shift in training data is welcome, but implementation will be criticism. “I’m glad to see OpenAI getting involved in this topic. Whether or not this will help artists will depend on the details, which have not yet been provided,” he states. The first important question on his mind: Is this simply an opt-out tool that allows OpenAI to continue using data without permission unless the content owner requests opt-out? Or will it represent a bigger change in the way OpenAI does business? OpenAI did not immediately respond to a request for comment.
Newton-Rex is also curious if OpenAI will allow other companies to use its Media Manager so that artists can indicate their preferences to multiple AI developers at once. “If not, you’ll just add more complexity to an already complex opt-out environment,” says Newton-Rex, a former executive at Stability AI, developer of the Stable Diffusion imager.
OpenAI is not the first to look for ways for artists and other content creators to indicate their preferences about the use of their work and personal data for AI projects. Other technology companies, from Adobe to Tumblr, also offer opt-out tools for data collection and machine learning. The startup Spawning launched a registry called Don’t train It’s been almost two years and creators have already added their preferences for 1.5 billion works.
Jordan Meyer, CEO of Spawning, says the company is not working with OpenAI on its Media Manager project, but is open to doing so. “If OpenAI can make it easier to sign up or honor opt-outs, we would be happy to add their work to our suite,” he says.