OpenAI partners with Hearst, now has more content to train AI

According to an exclusive report by Axios on Monday, OpenAI has inked a content partnership with Hearst, one of the nation’s largest newspaper and magazine conglomerates. The deal adds Hearst to the list of publishers collaborating with OpenAI, which already includes Condé Nast and Dotdash Meredith—competitors in the magazine industry.

Magazine giants’ content now under OpenAI’s belt

The partnership will allow OpenAI to incorporate content from over 40 of Hearst‘s local newspapers, including notable publications like the Houston Chronicle and San Francisco Chronicle. It will also include material from more than 20 of Hearst’s magazine brands, such as Esquire, Cosmopolitan, and Women’s Health. This content will be integrated into OpenAI’s products, including ChatGPT, providing more depth and diversity to the AI’s responses.

As with previous publisher agreements, OpenAI has confirmed that content used to generate responses in ChatGPT will come with “appropriate citations and direct links” to the original Hearst sources. While financial specifics of the deal weren’t disclosed, it’s speculated that Hearst is receiving substantial compensation, potentially millions, for licensing its intellectual property. Additionally, OpenAI confirmed that Hearst would receive credits to leverage its AI technology.

The partnership will allow OpenAI to incorporate content from over 40 of Hearst’s local newspapers (Image credit)

Jeff Johnson, president of Hearst Newspapers, noted, “This agreement allows the trustworthy and curated content created by Hearst Newspapers’ award-winning journalists to be part of OpenAI’s products like ChatGPT—creating more timely and relevant results.” Debi Chirichella, president of Hearst Magazines, added, “Our partnership with OpenAI will help us evolve the future of magazine content.”

Axios reports that these partnerships are moving away from providing broad training data for large language models (LLMs) toward more targeted use cases where news publishers maintain greater leverage. The approach now favors Retrieval Augmented Generation (RAG), a process where AI models access smaller, vetted datasets in real-time to deliver more accurate and relevant responses, especially for queries related to news or cultural events. This is the same strategy OpenAI is employing through its partnerships, including this recent agreement with Hearst.

Featured image credit: Kerem Gülen/Ideogram