OpenAI Introduces Media Manager Tool to Empower Creators’ AI Content Control
OpenAI is developing a new tool, Media Manager, that gives creators greater control over the use of their content in generative AI.
This new tool, Media Manager, will give the content creators and content owners the ability to define what kind of content can be used by the company, and under which conditions it should be included or excluded from AI research and training.
According to Open AI, this tool is expected to be implemented by 2025, as it is partnering with creators, content owners, and regulators, with the aim to develop a benchmark, probably through the steering committee they have recently joined.
In a blogpost, the AI company stated that “this will require cutting-edge machine learning research to build a first-ever tool of its kind to help us identify copyrighted text, images, audio and video across multiple sources and reflect creator preferences.”
It seems that this new tool, regardless of its form, is OpenAI’s answer to all the criticism it faced in relation to IP infringement. For instance, the company has encountered several lawsuits from reputable newspapers claiming that it has used its articles without their compensating them, such as The New York Times who claimed that OpenAI and Microsoft used its copyrighted material to train its AI models.
AI models, such as ChatGPT, are usually trained on a vast number of public sources and datasets. As for OpenAI and other supporters of generative AI believe that their practices are under fair use, others do not agree.
To this end, and to alleviate concerns and protect itself from being legally sued, OpenAI is taking actions to reach a consent with content creators.
Last year, the company introduced a system that allows artists to withdraw and remove their content from the datasets that were used to train image generating models. It also allowed the website owners to determine through the robots.txt protocol whether their content can be used for AI model training.
OpenAI is also engaging in negotiations for licensing agreements with significant content providers, including prominent news agencies, stock media libraries, and Q&A platforms such as Stack Overflow.
On the other hand, some artists have criticized the opt-out feature for images as a burdensome, as it requires submitting each copy of each image individually with descriptions for removal. The company reportedly pays the creators a minimal amount for content licensing.
From its part, OpenAI stated in the blogpost that the provided solutions do not tackle situations where creators’ works are quoted, revised, or reposted on platforms beyond their control.
Inside Telecom provides you with an extensive list of content covering all aspects of the tech industry. Keep an eye on our Intelligent Tech sections to stay informed and up-to-date with our daily articles.