OpenAI has decided to stop using customer data for training its models in response to criticism

OpenAI has changed its developer policy to reflect the criticisms of both developers and users.

OpenAI has announced that, starting today, it will not use any data sent through its API to \”improve services,\” including AI model-training, unless the customer or organization gives consent. The company will also implement a 30-day policy on data retention for API users, with the option of a stricter retention period \”depending upon user needs,\” as well as simplifying terms and data ownership so that it is clear users own both the input and the output from the models.

Greg Brockman is the president and chair of OpenAI. He says that these changes have always existed. OpenAI API users are the owners of input and output data. This includes text, images, and other types of data. He says that the new legal issues surrounding generative AI, as well as customer feedback, led to a rewrite of the terms of services.

Source:

Addressing criticism, OpenAI will no longer use customer data to train its models by default