OpenAI’s new Media Manager tool for creators

From BGR: 2024-05-08 20:41:00

OpenAI is facing scrutiny for using copyrighted content in training its AI models, prompting lawsuits and government investigations. Basic privacy protections were lacking, but the company plans to launch Media Manager, allowing creators to exclude their content from training. The tool is set to be ready by 2025.

The Media Manager tool aims to identify and exclude copyrighted text, images, audio, and video from training AI models like ChatGPT. OpenAI is collaborating with creators, content owners, and regulators to develop this feature for protecting intellectual property rights.

OpenAI asserts good intentions in developing AI products that benefit users. It clarifies the model training process and emphasizes that data used for training models is not retained. Models like ChatGPT do not store data and are constantly updated with new datasets for training.

OpenAI ensures that AI models like ChatGPT should not regurgitate content, attributing any repetition as a failure in the training process. The company is dedicated to diversity in training data, sourcing information from publicly available data and partnerships, excluding content with paywalls or personal information.

Users of ChatGPT can opt out of training the chatbot to protect their data privacy. This option is available for different versions of ChatGPT, including Team, Enterprise, API Platform, Free, and Plus. OpenAI emphasizes the importance of user privacy and control over data usage in AI training processes.



Read more at BGR: OpenAI’s new Media Manager tool for creators