OpenAI has reportedly announced a "first of its kind" partnership with Axel Springer, the Berlin-based publisher of magazines like Politico and Business Insider, to incorporate recent articles into OpenAI's popular AI-powered chatbot ChatGPT and train its generative AI models on the publisher's content.
Through this cooperation, Axel Springer's media brands, including POLITICO, BUSINESS INSIDER, and the European domains BILD and WELT, will reportedly provide summaries of selected global news material to ChatGPT customers worldwide, including otherwise paid content. For openness and further information, OpenAI states that ChatGPT will provide attribution and links to the whole article in response to user inquiries.
Reuters adds that Tom Rubin, head of intellectual property and content at OpenAI, said that the summaries will be available on ChatGPT as soon as the item is published, making breaking news a part of the user experience. Rubin adds that the Axel Springer material will debut in the first quarter 2024.
According to a person acquainted with the agreement, Reuters adds that the material will have a "favorable position" in ChatGPT search results, to support Axel Springer brands by increasing traffic and subscription income.
OpenAI's Journalistic AI Deals
TTech Crunch adds that this is only the second partnership OpenAI has had after it recently announced that it would license portions of The Associated Press archives for training models.
According to Reuters, The Associated Press and OpenAI reached a contract in July wherein the AP licenses a portion of its news story library to the Microsoft-backed software business. The deal's financial terms were not made public, but the AP will have access to OpenAI's technology and product knowledge.
Brad Lightcap, COO of OpenAI, lauded the newest partnership, adding that Through OpenAI's capabilities, the cooperation with Axel Springer will help provide consumers with new ways to receive high-quality, real-time news information.
According to the COO, they are steadfastly dedicated to collaborating with publishers and artists globally and ensuring they have access to cutting-edge AI technologies and novel income streams.
The partnership will reportedly see OpenAI paying Axel Springer in an amount and at an arbitrary frequency. The agreement is good for several years, and according to Axel Springer, it will help the outlet's current AI-driven projects "that build upon OpenAI's technology," even though neither party promises exclusivity.
Copyrighted Material as AI Training Data
Looking forward, there could reportedly be more partnerships with similar organizations. Robert Thomson, the CEO of News Corp., stated in November that the business was in discussions to reach agreements about using its material for generative artificial intelligence.
OpenAI's deal came after Meta and other technological firms were hit with copyright infringement cases throughout the year.
Meta, in particular, is being accused of being aware of the legal consequences of training its AI model "Llama" with copyrighted books from the controversial dataset "The Pile," including the "Books3" collection containing copyrighted books and articles from the past 20 years without the authors' permission, as previously reported by Tech Times.
