OpenAI’s legal tussle with The New York Times may be unresolved, but the organization is pressing forward with significant deals with other publishers, including major players in France and Spain.
In a recent announcement, OpenAI revealed contracts with Le Monde and Prisa Media, facilitating the integration of French and Spanish news content into OpenAI’s ChatGPT chatbot. This partnership will not only provide users with access to current events coverage from renowned brands such as El País, Cinco Días, As, and El Huffpost but also contribute to the expansion of OpenAI’s training data repository.
Over the next few months, ChatGPT users will engage with relevant news content through select summaries with attribution and enhanced links to original articles, enhancing their ability to access additional information or related articles from these news sites.
OpenAI underscores its commitment to continually improving ChatGPT while supporting the critical role of the news industry in delivering real-time, authoritative information to users.
In partnership with Le Monde and Prisa Media, our goal is to enable ChatGPT users around the world to connect with the news in new ways that are interactive and insightful.
Brad Lightcap, COO of OpenAI
OpenAI’s roster of licensing deals now includes several notable names, including Shutterstock, The Associated Press, Axel Springer, Le Monde, and Prisa Media. While the exact financial terms remain undisclosed, estimates suggest substantial annual investments ranging from $4 million to $20 million for news licensing alone, based on previous reporting.
This level of investment underscores OpenAI’s formidable financial position, with a war chest exceeding $11 billion and annualized revenue surpassing $2 billion. However, it also raises questions about potential barriers to entry for AI competitors seeking similar licensing agreements.
Hunter Walk, a partner at Homebrew, cautions that exorbitant licensing costs could stifle innovation and limit competition in the AI landscape. The proliferation of AI models trained on copyrighted data, without formal agreements, further complicates the regulatory landscape.
The debate around licensing costs prompts consideration of regulatory interventions, such as a proposed “safe harbor” to protect AI vendors, startups, and researchers from legal liability, provided they adhere to transparency and ethical standards.
However, regulatory efforts in this direction have faced challenges, raising broader questions about the role of copyright in AI development.
In the context of a news industry already grappling with instability, fair compensation for publishers is paramount. Yet, there is a need to balance this with ensuring equitable access to training data for challengers to AI incumbents and academic researchers. While solutions remain elusive, exploring avenues for grants, larger VC investments, and potential carve-outs could offer a path forward.