OpenAI can now access Financial Times articles to train AI


Financial Times, a publication read by senior financial executives with a $75 per month price for full internet access, is allowing OpenAI to use its articles to train an AI chatbot.

OpenAI's ChatGPT will soon answer relevant questions with a summary and quotes taken directly from the FT pages, for a deal announced Monday. The chatbot will link back to the full FT articles it refers to.

FT Group CEO John Ridding said that “it's right, of course, that AI platforms pay publishers for the use of their material” and that OpenAI “understands the importance of transparency, attribution and compensation – all essential to us.”

Financial details of the partnership, as well as when the FT articles will be incorporated into OpenAI products, were not disclosed.

John Ridding, CEO, Financial Times. Photo by Sportsfile/Corbis/Sportsfile via Getty Images

In addition to the FT, OpenAI signed a similar deal with Axel Springer, the publisher behind Politico, Business Insider and Bild, in Decemberwhich means ChatGPT can soon cite the company's content.

Under the agreement, ChatGPT will aggregate articles from Axel Springer's brands as they are published in real time and link to full articles, potentially driving more traffic to the publisher's sites.

Connected: The authors are suing OpenAI because ChatGPT is too 'accurate' – here's what that means

Newspaper publishers are suing OpenAI

While the FT and Axel Springer are examples of two organizations that chose to work with OpenAI, other publications are taking legal action against ChatGPT's creator alleging copyright infringement.

Eight newspapers owned by hedge fund Alden Global Capital, including the New York Daily News, Chicago Tribune and Orlando Sentinel. ignorant OpenAI and Microsoft on Tuesday.

The publications claimed that both ChatGPT and Copilot, AI chatbots provided by OpenAI and Microsoft respectively, can repeat long snippets of articles behind paywalls with the right request.

This may indicate that chatbots have been trained on copyrighted material used without permission or payment.

Connected: Sarah Silverman is suing OpenAI and Meta for 'copyright infringement', using her works to train AI models

Microsoft and OpenAI are “taking the work of publishers with impunity and using publisher journalism to create GenAI products that undermine publishers' core businesses by rebroadcasting 'their content'” — in some cases verbatim from publisher websites with paywalls – in them the readers”, the complaint DECLARING.

AI robots can too they hallucinateor misrepresents information and attributes that information to publications, according to the complaint.

Other news organizations have made similar claims against OpenAI and Microsoft in previous legal actions. New York Times filed a lawsuit in December alleging that the two tech giants used millions of its articles “to create products that replace The Times and steal its audience.”

News sites The Intercept, AlterNet and Raw Story filed their own lawsuits February.

Connected: Elon Musk Sues ChatGPT-Maker OpenAI



Source link

Leave a Reply

Your email address will not be published. Required fields are marked *