Comprehensive Collaboration: Media Giant and OpenAI Seek Win-Win Cooperation

[baoshi.rao]

On December 13, OpenAI announced a strategic partnership with Axel Springer, Europe's largest digital publishing company. This marks the first comprehensive strategic collaboration between a media giant and an AI company in terms of copyright, technology, and business models, following repeated protests by media alliances against AI manufacturers.

This article outlines the significant implications of this collaboration for the news industry and highlights key aspects of the cooperation agreement, shared here for the reference of domestic industry peers.

On December 13th, OpenAI announced a strategic partnership with Axel Springer, Europe's largest digital publishing company. Neither company disclosed financial terms in their press releases, but industry observers expect the agreement to generate significant revenue for Axel Springer. Sources familiar with the matter revealed that OpenAI had previously discussed using article word count as a metric for determining content compensation.

Under this partnership agreement, ChatGPT users will receive summaries of "selected" articles from Axel Springer's publications—including those typically behind paywalls. These summaries will include attribution and links to the full articles.

In return, Axel Springer will receive payments from OpenAI of unspecified amounts and frequencies. The agreement is valid for several years, and while it does not require exclusive cooperation between the two parties, Axel Springer has stated that it will support existing AI-driven projects "based on OpenAI technology."

Analysts believe this is a significant milestone, representing how media companies can receive compensation for their content from AI tools. This could potentially become a prototype for a new business model, where news organizations sign similar long-term agreements with companies providing AI services, such as Google and Microsoft.

Headquartered in Germany, Axel Springer is a company with numerous well-known media brands, including POLITICO, BUSINESS INSIDER, BILD, and WELT. This collaboration makes it the second publisher worldwide to partner with OpenAI.

In July this year, the Associated Press (AP) reached an agreement with OpenAI, authorizing the use of part of AP's news archive to explore the application of generative AI in the field of journalism for model training. However, it is reported that this collaboration does not authorize the use of AP content in ChatGPT's user outputs.

The rapid rise of generative AI has sparked anxiety across the publishing industry, with concerns in journalism about issues such as traffic diversion. According to a recent survey, hundreds of news organizations are now using code to prevent OpenAI, Google, and other companies from scanning their websites for training data.

In August of this year, several media organizations, including Getty Images, the Associated Press, the National Press Photographers Association, and the Writers Guild, released an open letter calling for increased transparency and copyright protection in the AI sector. In the letter, the signatories urged policymakers to consider regulations that mandate transparency and allow media companies to negotiate with AI model operators, among other recommendations.

The letter stated: 'Current practices undermine the core business models of the media industry, which rely on readers and viewers (e.g., subscriptions), licensing, and advertising. Beyond violating copyright laws, the result is a significant reduction in media diversity, weakening the financial capacity of companies to invest in news coverage, and further limiting public access to high-quality and trustworthy information.'

While journalism resists AI companies, the development of large models also faces bottlenecks. OpenAI previously collected large amounts of data from platforms such as Wikipedia, blogs, Twitter, and Reddit to train AI models, but these sources are said to be increasingly unable to meet the demands of large model training. AI models now need to find reliable and continuously updated data sources.

A research report released in October this year by the News Media Alliance, a trade organization comprising over 2,200 publishers, indicates that datasets used to train popular AI models rely "significantly" more on publisher content compared to general web content, with proportions ranging from over five times to nearly 100 times.

To make model content richer and more accurate, AI large model vendors signing agreements with authoritative media holds significant strategic value. The specific collaboration details are as follows:

1. ChatGPT users will gain access to curated global news summaries and premium content from Axel Springer's media brands.

ChatGPT users will receive customized global news summaries from Axel Springer's media brands, featuring exciting headlines, in-depth reports, and engaging feature content. Even more enticing, users will have access to exclusive paid content, including carefully curated special reports, in-depth analyses, and exclusive interviews. This collaboration will open a window for ChatGPT users to delve into the essence of Axel Springer's media offerings, providing a richer and more comprehensive news experience.

2. ChatGPT responses to user queries will include the source and full link to the articles.

ChatGPT's responses will comprehensively present the sources of articles and provide complete links, ensuring users can conveniently and quickly access relevant information. This design not only helps users gain an in-depth understanding of the sources of responses but also offers direct access to the original text to meet users' needs for more detailed information. We are committed to providing users with the most comprehensive and transparent information, enabling them to delve into the knowledge and insights behind the answers while obtaining them.

3. Axel Springer's media brand content will be used for training OpenAI's large language models.

OpenAI's large language models will fully utilize Axel Springer's media brand content for training. This collaboration will provide the AI language models with a rich and diverse corpus, covering content from Axel Springer's well-known media brands. By integrating these abundant information resources, the aim is to further enhance the language models' comprehension, text generation capabilities, and breadth of knowledge, offering users more accurate and comprehensive responses.

4. OpenAI will provide technical support for Axel Springer's AI projects.

OpenAI will provide technical support for AI projects to Axel Springer, ensuring its top-tier technology and innovation in the artificial intelligence field are utilized to their fullest potential. Both parties will jointly explore how to best integrate AI technologies to enhance the efficiency and innovation of Axel Springer's media brands. This technical support collaboration will help Axel Springer better understand and leverage the potential of AI, driving continuous development in the digital media sector.

5. OpenAI will pay Axel Springer for content used in training large language models.

Although the specific financial terms were not mentioned in the statement, according to Reuters, OpenAI will compensate Axel Springer for the content used in training large language models. Additionally, this collaboration is expected to bring more traffic and subscription revenue to Axel Springer.

For the news publishing industry, which possesses rich and high-quality copyrighted language data, companies like OpenAI are seeking win-win scenarios. Brad Lightcap, OpenAI's Chief Operating Officer, stated that the partnership with Springer will help provide people with new ways to access high-quality real-time news content through AI tools. "We are firmly committed to working with publishers and creators worldwide to ensure they benefit from advanced AI technologies and new revenue models."

OpenAI stated: "We are thrilled to establish this global partnership between Axel Springer and OpenAI—the first of its kind. We aim to explore the opportunities AI presents for journalism—enhancing its quality, social relevance, and business models to new heights."