The Financial Times (FT) this morning announced a deal with OpenAI, the company behind generative AI tools including ChatGPT, Dall-E and Sora, allowing the tech firm to use FT content to train its AI models.
The announcement makes FT the latest in a string of a publishers to strike such agreements with the Microsoft-backed startup, following similar deals with Axel Springer, Le Monde and Associated Press. As well as licensing the FT’s material to OpenAI for developing GenAI tools, the partnership will also allow ChatGPT to respond to questions with summaries from FT articles.
“Our partnership and ongoing dialogue with the Financial Times is about finding creative and productive ways for AI to empower news organisations and journalists, and enrich the ChatGPT experience with real-time, world-class journalism for millions of people around the world,” said OpenAI COO Brad Lightcap.
“Deals with the devil”
Deals of this nature have sparked controversy in the industry, raising doubt over publishers receiving fair renumeration for their content, particularly smaller news publications without the means to secure favourable licensing terms.
Meanwhile The New York Times is suing OpenAI, claiming that the company has scraped millions of its articles to train ChatGPT, which it says has regurgitated extracts of Times articles verbatim in its responses.
But others see these agreements as something of a necessary evil, not unlike previous deals with social media companies which once provided news publishers a vital source of traffic. Last week, Axel Springer CEO Matthias Dopfner was questioned on the company’s “deal with the devil”, responding that given the threat large language models (LLM) pose to journalism, it was better to set a precedent whereby publishers are paid by AI companies for access to their content.
And this appears to be the thinking from the FT, arguing that the deal protects the sustainability of its business. “It’s right, of course, that AI platforms pay publishers for the use of their material,” said FT chief executive John Ridding. “OpenAI understands the importance of transparency, attribution, and compensation – all essential for us. At the same time, it’s clearly in the interests of users that these products contain reliable sources.”
The company notes that FT content reproduced by ChatGPT will link back to FT.com, meaning the chatbot’s 100 million users can access the FT’s reporting, while also having a route back to the original source material.
In at the ground floor
The FT is no stranger to AI, noting its readiness to embrace the technology at the earliest signs of disruption. The publisher subscribes to ChatGPT Enterprise, OpenAI’s business offering, granting all employees access to the tool. And the company has recently created an AI editor role, to both report on the topic and help educate the editorial team.
Again, the extent to which newsrooms are harnessing AI for editorial purposes varies between publishers. Reach, owner of the Mirror and Daily Express, recently launched its own in-house AI tool that repurposes the publisher’s content across its different brands. But a recent survey by the Association of Online Publishers (AOP) suggested that one-quarter of UK publishers have editorial policies in place prohibiting the use of generative AI.
