OpenAI and The Associated Press (AP) have announced a licensing deal that grants OpenAI access to AP’s extensive archive of news stories. The agreement also allows AP to leverage OpenAI’s technology and product expertise, according to a joint statement by the two organizations. The financial terms of the deal were not disclosed.
AI Companies Depend on Text Archives
AI companies, including OpenAI, rely on large collections of written works, such as books, news articles, and social media content, to enhance their AI systems, often referred to as large language models. OpenAI’s release of ChatGPT last year led to a surge in “generative AI” products capable of producing new text passages, images, and other media. However, concerns have been raised regarding the potential for these tools to generate false information that may go unnoticed due to their strong command of human language grammar. Furthermore, questions have emerged about the extent to which news organizations and other content creators should be compensated for their contributions to training AI models.
Authors and News Organizations Seek Compensation
Authors, alongside news organizations, have been advocating for compensation for their works used in training AI systems. Over 4,000 writers, including Nora Roberts, Margaret Atwood, Louise Erdrich, and Jodi Picoult, recently signed a letter accusing OpenAI, Google, Microsoft, Meta, and other AI developers of exploitative practices. They claim that chatbots developed by these companies “mimic and regurgitate” their language, style, and ideas. Some novelists and comedian Sarah Silverman have even filed copyright infringement lawsuits against OpenAI. The AP deal signals OpenAI’s recognition of the value of intellectual property and the importance of fact-based, nonpartisan news content, as stated by Kristin Heitmann, AP’s Senior Vice President and Chief Revenue Officer.
 strike a licensing deal, granting OpenAI access to AP's news archive, while AP benefits from OpenAI's technology expertise.){kind=link}