OpenAI, creator of viral AI chatbot ChatGPT, has struck another news licensing deal in Europe, adding London's Financial Times to a growing list of publishers it pays for access to content.
As with OpenAI's previous publisher license agreements, the financial terms of the agreement are not made public.
The latest agreement seems a little more comfortable than other recent tie-ups with OpenAI publishers, such as with the German giant. Axel Springer or with the P.A., Le Monde and Prisa Media in France and Spain respectively – with both men calling the deal a “strategic partnership and licensing agreement”. (Although the CEO of Le Monde also mentioned the “partnership” announced with OpenAI in March as a “strategic movement”.)
However, we understand that this is a non-exclusive licensing agreement – and OpenAI does not take any ownership interest in the FT Group.
Regarding content licensing, both partners said the agreement covers OpenAI's use of FT content for training AI models and, where applicable, for display in responses. Generative AI produced by tools like ChatGPT, which is very similar to its other deals with publishers.
The strategic element appears to focus on the FT strengthening its understanding of generative AI, particularly as a content discovery tool, and what is being presented as a collaboration aimed at developing “new AI products and features for FT readers” – suggesting the news publisher is keen to expand its use of AI technology more generally.
“Through this partnership, ChatGPT users will be able to see some attributed summaries, quotes and rich links to FT journalism in response to relevant queries,” the FT wrote.n / A Press release.
The publisher also noted that it became a customer of OpenAI's ChatGPT Enterprise product earlier this year. He goes on to suggest that he wants to explore ways to deepen its use of AI, while expressing caution about the reliability of automated outputs and potential risks to reader trust.
“This is an important agreement in several respects,” wrote FT Group CEO John Ridding in a statement. “This recognizes the value of our award-winning journalism and will give us a first look at how content appears using AI.”
He continued: “As well as the benefits for FT, there are wider implications for the industry. It is of course true that AI platforms pay publishers for the use of their material. OpenAI understands the importance of transparency, attribution and compensation, all of which are essential to us. At the same time, it is clearly in the interests of users that these products contain reliable sources.”
Large language models (LLMs) such as OpenAI's GPT, which powers the ChatGPT chatbot, are known for their ability to fabricate information or “hallucinate.” This is the opposite of journalism, where journalists strive to ensure that the information they provide is as accurate as possible.
It's no surprise, then, that OpenAI's first moves toward licensing content for training models focused on journalism. The AI giant may hope that this will help it solve the problem of “hallucinations”. (A line in the PR suggests that the partnership will “help improve the usefulness of (OpenAI) models by learning from FT journalism. “)
There is, however, another major motivating factor at play here: legal copyright liability.
Last December The New York Times announced that it was suing OpenAI, alleging that its copyrighted content was used by the AI giant to train models without a license. OpenAI disputes that But one way to avoid the risk of further lawsuits from news publishers, whose content has likely been removed from the public Internet (or otherwise harvested) to fuel the development of LLMs, is to pay publishers for the use of their copyrighted content.
For their part, publishers have every chance of making real money through content licensing.
OpenAI told TechCrunch that it has “about a dozen” signed (or “imminent”) deals with publishers, adding that “many” more are in the works.
Publishers could also potentially acquire some readers, for example if ChatGPT users choose to click on quotes that link to their content. However, generative AI could also cannibalize search engine usage over time, driving traffic away from news publishers' sites. If this type of disruption occurs, some news publishers may feel a strategic advantage in developing closer relationships with companies like OpenAI.
Getting involved in Big AI also comes with reputational pitfalls for publishers.
Technology publisher CNET, which rushed to adopt generative AI as a content production tool last year — without making the use of the technology very clear to readers – further damaged his reputation when Futurism journalists discovered dozens of errors in automatically written articles he had published.
The FT has a well-established reputation for producing quality journalism. It will therefore certainly be interesting to see how it further integrates generative AI into its products and/or its editorial processes.
Last month, announcement a GenAI tool for subscribers – which is essentially offering a natural language search option over two decades of FT content (so it is, fundamentally, value-add aimed at generating subscriptions for journalism produced by man).
Additionally, in Europe, legal uncertainty clouds the use of tools like ChatGPT. a raft of privacy law concerns.