OpenAI hasn’t trained its AI big-language models like GPT on paid customer data “for a while,” CEO Sam Altman told CNBC on Friday.
“The customers clearly don’t want us to train on their data, so we’ve changed our plan: We won’t do it,” Altman told CNBC’s Andrew Ross Sorkin.
OpenAI’s terms of service were quietly updated on March 1, records from the Internet Archive’s Wayback Machine show. “We don’t train on any API data, we haven’t for a while,” Altman told CNBC. APIs, or application programming interfaces, are frameworks that allow customers to plug directly into OpenAI’s software.
OpenAI’s commercial customers, including Microsoft, sales force And SnapchatThere is more potential to leverage OpenAI’s API capabilities.
But OpenAI’s new privacy and data protection is limited only to customers who use the company’s API services. “We may access content from services other than through our APIs,” the company’s updated terms of use note. This can, for example, include text that employees enter into the wildly popular chatbot ChatGPT. Amazon Allegedly Recently employees were warned not to share confidential information with ChatGPT for fear that it could appear in replies.
The change comes at a time when the industry is grappling with the prospect of a larger language model replacing man-made content.
For example, the Writers Guild of America started strike On Tuesday, talks between the guild and the film studio broke down. The guild was pushing for limits on OpenAI’s use of ChatGPT for creating or rewriting scripts.
Executives are equally concerned about the impact of ChatGPT and similar programs on their intellectual property. Barry Diller, the entertainment mogul and president of the IAC, has suggested Media companies can take their issues to the courts and potentially sue AI companies over the use of creative content.
