Financial Markets


In a move to both expand the intellectual scope of its language model and safeguard itself from copyright violations, OpenAI, the artificial intelligence research lab, has announced the signing of licensing agreements with two major media companies - The Atlantic and Vox Media. With this move, OpenAI is strategically navigating the intersection of AI and the copyright landscape, a terrain littered with potential legal pitfalls. Adding fuel to the fire, this new AI paradigm is simultaneously accelerating the digital transformation across the media industry and shaking up the tenets of journalism as we know it.

The new agreements allow OpenAI to utilize the content of these media companies to feed its AI models, and share within its ChatGPT – an advanced AI chatbot. Previously, OpenAI has notably inked similar deals with News Corp, Axel Springer, DotDash Meredith, the Financial Times, and The Associated Press. The price tag of these agreements depends on the expanse of publications involved, with News Corp’s deal estimated to be a hefty $250 million over five years.

This partnership doesn’t stop at filling OpenAI's data banks; it also grants provisions for showcasing publisher content within ChatGPT, with full attributions provided when the material is regarded. This continues OpenAI’s focus on partnering with media organizations, not only for augmenting its training data sets but also to circumvent copyright lawsuits, most famously the ongoing lawsuit by The New York Times.

Both The Atlantic and Vox Media are not merely providing fodder for AI training but are also deploying OpenAI's technology for their ventures. Vox Media is planning to bolster one of its products using AI, expanding its ad data platform in the process. On the other hand, The Atlantic is ambitiously developing a microsite, experimenting with AI tools to improve journalism and reader engagement.

The potential impact on the future of AI-engineered journalism and the media industry at large is profound. Artificial intelligence algorithms, when trained with a plethora of articles, could possibly generate news reports without human intervention or offer novel analytics and insights for journalists. Inversely, the concern of AI-generated “deepfake” news increases. In the copyright domain, this might set a precedent for more technology companies to partner with content providers and explore legal and innovative ways to train their AI models.

The OpenAI licensing agreements may serve as a roadmap for AI companies grappling with copyright issues while showing that AI and journalism needn't be adversaries but could collaboratively determine the trajectory of future media. This shift presents a future where AI doesn’t replace journalists but aids them, growing a symbiotic relationship that redefines the rules of journalism.