OpenAI, an artificial intelligence research lab led by CEO Sam Altman, has announced that it won’t train on customer data anymore- a marked departure from its original plans. This follows a demand from customers who want the company to safeguard data. Altman told CNBC’s Andrew Ross Sorkin that customers “clearly want [them] not to train on their data,” and so the company has changed their plans accordingly.
OpenAI’s terms of service have been quietly updated on March 1, and now they claim they won’t train on any data from Application Programming Interfaces (APIs). This feature is increasingly useful for businesses, including Microsoft, Salesforce, and Snapchat.
The update to OpenAI’s policy has major implications in the creative industry. Writers Guild of America had been pushing for limitations on OpenAI’s large-language models, like ChatGPT, from script generating. This eventually led to them striking. Other media moguls, including IAC chairman Barry Diller, have noted the threat of such technology to intellectual property; some have gone as far as suggesting to take the issue to the court.
OpenAI was founded in 2015 as a non-profit research lab, and has quickly become one of the most renowned AI development companies, working on machine learning, research and software. Since then, it’s been focusing on developing AI that can think, reason, and solve problems. Under the direction of Altman, OpenAI has amassed an impressive list of accomplishments, including creating machine learning models that surpassed humans in complex tasks like language understanding and playing games.
