OpenAI, embroiled in legal disputes with The New York Times over data usage for AI training, is moving forward with licensing deals with other publishers, including some of France’s and Spain’s largest news organizations.
The tech giant recently announced agreements with Le Monde and Prisa Media to integrate French and Spanish news content into its ChatGPT chatbot. This partnership will allow ChatGPT users to access current events coverage from renowned brands such as El País, Cinco Días, As, and El Huffpost, enhancing the chatbot’s capabilities and enriching user experience.
While the financial terms of these deals remain undisclosed, estimations suggest significant sums, with reports indicating that OpenAI is offering between $1 million and $5 million annually to publishers for access to archives to train its AI models. Considering the scale of OpenAI’s revenue and resources, these figures underscore the company’s commitment to securing valuable training data.
However, the proliferation of such licensing agreements raises ethical concerns and potential barriers to entry for competitors in the AI space. Hunter Walk, a partner at Homebrew and co-founder of Screendoor, warns that exorbitant licensing fees could stifle innovation and create monopolistic advantages for large tech companies like Google and OpenAI.
The debate over the necessity of licensing for AI experimentation is ongoing. While some argue that it should be an essential cost of doing business, others advocate for regulatory interventions to ensure fair access to training data and protect smaller players in the industry.
Amid these discussions, the role of publishers in shaping the future of AI remains pivotal. While they deserve fair compensation for their content, there is also a need to facilitate broader access to data for AI research and development. Finding a balance between these competing interests is crucial to fostering innovation while upholding ethical standards and ensuring a level playing field in the AI landscape.
