Reports indicate that Google is negotiating with publishers regarding AI licensing agreements amid increasing tensions between the media and AI sectors.
Bloomberg has revealed that Google intends to launch a “pilot initiative initially involving around 20 national news organizations,” which would enable these participants to license their content for Google’s AI applications. While specifics are minimal, the strategy appears similar to that of OpenAI, which has previously established licensing agreements with prominent publishers such as Hearst, Condé Nast, Vox Media, The Atlantic, and News Corp. Perplexity holds the second position in the tally of agreements with publishers.
In this context, numerous publishers have indicated that AI tools like ChatGPT, Google AI Overviews, and Google AI Mode have resulted in diminished traffic. A recent report by the Wall Street Journal characterized the scenario as “AI Armageddon” for digital news publishers, asserting that they are being “overwhelmed” by Google’s AI search features. An article in The Economist was even more direct, proclaiming that “AI is destroying the web,” accompanied by an image of a gravestone.
Google has already formed an AI licensing collaboration with The Associated Press to deliver real-time news updates through its Gemini model. Additionally, it has a $60 million licensing agreement with Reddit. Nevertheless, this prospective pilot could greatly widen this initiative. “We have indicated that we are investigating and testing new partnership types and product experiences, but we are not disclosing details about specific plans or discussions at this moment,” stated a spokesperson for Google.
Media firms face a challenging decision: Oppose AI firms or collaborate with them
The publishing sector is split on how to approach the utilization of its content for AI model training. AI companies’ bots extract data from the internet for valuable training information, which is then used to influence chatbot replies.
Some publishers and authors have accused firms of copyright violations for utilizing this content without consent or recompense. Currently, The New York Times is embroiled in a lawsuit against OpenAI and Microsoft for this reason, among numerous similar legal actions. (Mashable’s parent company Ziff Davis is also involved in a lawsuit against OpenAI for copyright infringement.)
Conversely, other publishers have opted to license their content, pointing to new avenues for readers to uncover their narratives. Although the specifics of licensing arrangements with OpenAI haven’t been made public, reports suggest that Dotdash Meredith earns $16 million annually, while a report from The Information claims some publishers receive as little as $1 million yearly.
The assertions from tech companies that the use of scraped content falls under the protection of the fair use legal doctrine remain legally untested. Although Anthropic and Meta recently secured victories against authors based on the fair use argument, a pre-publication draft of a high-profile AI report from the U.S. Copyright Office largely supported copyright holders in the context of AI training. As courts evaluate particular fair use situations, the expanding AI licensing market may suggest that tech companies acknowledge the importance of collaborating with publishers for quality data.
Simultaneously, Google’s rollout of AI-generated summaries and AI Mode continues to diminish outbound traffic, according to various publisher reports. Instead of navigating to external sites from Google search results, users receive information from Google’s AI models directly on the search results page. On its AI site, Google states it is “collaborating with the ecosystem to investigate new partnership and value-exchange models.”
With the surge of generative AI altering the digital media environment, Google could play a pivotal role in shaping the future of online publishing.
Disclosure: Ziff Davis, the parent company of Mashable, initiated legal action against OpenAI in April, claiming it violated Ziff Davis copyrights in the training and functioning of its AI systems.