Welcome to the Publisher Newsletter by Media Voices: profiling the people and products powering publishing.

The FT’s Jon Slade on leveraging AI to build new lines of business

Jon Slade, recently appointed as CEO of The Financial Times, told the audience at The Definitive AI Forum for Media, Information & Events, that doubling down on original journalism and strong brand values is the answer to many of the challenges faced by today’s media businesses. However, he also sees opportunities for new lines of business powered by AI.

Jon Slade took over as CEO of The Financial Times from John Ridding in July 2025. With subscriptions accounting for more than half of the publisher’s revenue, the years before he took over will be remembered as the era of paid content for the business publisher.

Looking ahead to what may come to be seen as the era of generative AI, Jon sees clear challenges. For him, the answer lies in going back to first principles: investing in original journalism, and in telling the story of the FT brand.

“Neither of those are technology solutions,” he explained to Seedelta’s Chris Duncan on stage at the Definitive AI Forum. “But they are responses to a technology issue.”

However, that doesn’t mean Jon doesn’t see solutions and opportunities in technology. “Facing up to future challenges means leaning into technology,” he said. “Whether it’s experimenting with it for our own internal use, or developing new lines of business.”

AI-powered licensing

Licensing is high on the FT’s new business list, but not just the large-language deals that have dominated the industry over the last couple of years.

Two-and-a-half years after it signed its first LLM deal, the FT has a range of agreements in place. “They include different parts of our archive, from lots of it to not much,” Jon outlined. “Some of them include live feeds. Some do not. Some have a delay on them. Some don’t. Some allow for surfacing. Some don’t. There’s quite a spectrum.”

The one thing the large-language deals have in common is that Jon doesn’t see them as a long-term option. While he was broadly happy with the remuneration achieved, he said he didn’t believe licensing content to Big Tech to train their LLMs is a sustainable business model.

“If I’ve learned anything from my 20 odd years at the FT, the mantra is recurring revenue,” he explained. “I think it requires much more regulatory pressure to really make those people come to the party on a consistent basis.”

Jon is much more confident in direct-to-business licensing. “I think that’s a very exciting area for us to develop a much more sustainable business around,” he commented.

That is partly about control of content. Jon explained that a common requirement in every LLM deal is what he calls ‘The Big Red Button’. “If we don’t like what we see – assuming the transparency commitments are honoured – we push the button. It doesn’t take the salt out of the soup, but it stops archive access and stops surfacing. If we can’t get that, we don’t do the deal.”

In contrast, corporate licensing deals put the reins back in the FT’s hands. “There’s much more control for us in that business, which allows us to be much more confident about how it’s being used, how it’s being surfaced, and to set controls around usage. I think that’s a very exciting and sustainable business for us,” he said.

Jon explained that the FT already has an excellent licensing business built around ‘human licensing’: “We also call that subscriptions,” he laughed. He described how, alongside human subscriptions, the FT is expanding into the AI era with machine-reading licenses that will provide a data layer, metadata and archive access to corporate customers.

Unlocking value

AI has been a particularly powerful tool when it comes to unlocking the value in the FT’s archives. 

Jon described how AI had allowed the FT to reimagine its 3-million article archive using natural language processing and ‘some large language model magic’. He explained how, by looking at 30-years worth of archive content using AI, the FT could beat market predictions on what the Bank of England would do with interest rates 10 months out of 12. “We can predict US inflation better than the Fed,” he said.

Jon said this approach could also challenge traditional licensing models based on the article, seeing two potential approaches. In the first, the FT delivers a service that separates the signal from the noise on behalf of customers. “We do all of the work. You tell us what you want and we find the signal,” he explained.

In the second version the FT provides an intelligence feed that customers can manage according to their own needs. “We’ve got a pipe and we don’t know what you want to do with it. Here are the controls and what we will let you do with it, but we’ll plug it in and you find what solutions you need from it.”

Beyond simply being a smart example of AI technology in the newsroom, Jon said the use of AI could signal a shift in how the FT sees its journalism. “There’s intelligence that we can extract and it defines slightly differently what journalism could be in the future.”

The challenge will be for publishers to shift the perception of themselves as not just finding and reporting on stories, but interrogating content that they already have, possibly in tandem with third-party data sets.

“I think what journalism could be at the FT redefines us, potentially, not as a newspaper, but as an intelligence source. That puts us in a different operating theatre,” he concluded.

Listen to the full session in this week’s episode of The Publisher Podcast, available wherever you find podcasts.

More good reads: