10/30/2025 | News release | Distributed by Public on 10/30/2025 09:08
2025-10-30. At WAN-IFRA's recent Paris AI Forum, optimism defined the conversations. Rather than yielding to disruption, news executives and technologists spoke of agency, collaboration and a human-centered vision for the future of journalism.
by Teemu Henriksson [email protected] | October 30, 2025
Across the keynotes and panel discussions of Paris AI Forum, one message was clear: publishers are not powerless in shaping how AI transforms media.
Speakers from Reuters, The Independent, Les Échos, the BBC, and others presented their visions of how integrating AI helps journalism to streamline operations, empower journalists, and deepen trust with audiences.
Here are five themes that captured the state of the media-AI conversation in 2025.
"Publishers still have agency," said Christian Broughton, CEO of The Independent.
Rather than reacting defensively to technological change, Broughton urged his peers to seize AI as a creative catalyst.
"The best way to confront the future of AI in journalism is to use more AI, and not to consider yourself as passive in this process. Not to just find out what the tech business thinks you should do, or what the government thinks you should do, or how you can play defence."
He described how The Independent built Bulletin, an AI-powered news product that uses only the outlet's own content, with humans firmly in the loop. The project, developed in close collaboration with journalists, actually created jobs rather than eliminating them, with a new team in place running Bulletin's operation.
A similar approach guided iTromsø's work in Norway. The company's Head of AI, Lars Adrian Giskeexplained that, in the face of local competition, his newsroom created tools to automate data-mining tasks, which freed up time for reporting and fact-finding.
"Journalism is a human-driven mission for transparency, access to information, and telling stories that matter," Giske said.
"No agent system can do that, even if they can produce things that look like journalism."
Automation and efficiency were recurring themes, but speakers repeatedly emphasised that the human factor must remain at the heart of AI innovation, both in the newsroom and audience perception.
In France, Les Échos and Le Parisien showed how editorial transparency guides their experiments with AI-generated video and translation. Le Parisien's AI-augmented storytelling includes clear disclaimers, while its use of a cartoony visual style in video reconstructions ensures synthetic imagery is not mistaken for real footage.
For Olle Zachrison, the BBC's Head of News AI, transparency is non-negotiable. The broadcaster uses a task-tuned BBC LLM to adapt stories from local reporters and align them with their house style and format.
"Of course, an editor sub-edits and checks everything before publication," and the story is published with an AI disclaimer, he said.
The underlying message was that trust is built not just through what technology does, but how it is used: visibly, responsibly and with editorial accountability.
Beyond effective AI use cases, publishers discussed the need for collective defence against AI disruption, particularly in safeguarding intellectual property from large technology platforms.
David Buttlefrom DJB Strategies warned that publishers signing individual licensing deals risk making it harder for the content licensing market to form around collectively agreed market norms.
There have, of course, already been such deals, but Buttle didn't expect them to persist in their current form: "I still think there's a lot to play for, but the game is rapidly moving away from us as an industry," he said.
Meanwhile, Arnaud Aubron, Business Development Officer at Le Monde, and Martin Schori, Deputy Managing editor and Associate Publisher at Aftonbladet, shared the perspectives from their companies' early partnerships with OpenAI. They particularly highlighted the advantage of having access to AI-companies' engineers when building new AI-based products.
Elsewhere, Lucky Gunasekaraof Miso.ai called for "a NATO for news" - a coordinated approach to combat content scraping and unauthorised data use - and Josh Freemanof ProRata.ai presented shared licensing and attribution systems to ensure fair compensation.
Speakers also stressed that internal culture has a big influence on successful AI adoption.
Tina Rogers, Head of Delivery - Editorial AI and Automation at the Better Collective in Denmark, described a two-year process of trial, error and iteration in her sports media group. Her team worked on AI-assisted sports coverage to speed up content protection, learning how automation can be used safely with certain formats.
"If you are not already experimenting, start small and scale smart," she advised. "Prioritise people and communication over tech."
At Reuters, a similar process is happening but at a broader scale. Jane Barrett, the agency's Head of AI Strategy, said they measure AI success by speed, accuracy and productivity - but also crucially, "psychological safety" in creating environments where people truly can experiment with new tools.
"We want them to fail," Barrett said. "It means we tried something. It didn't work. So let's try it again, let's work it out. That's the way that we build the flexibility muscle."
Finally, many speakers discussed the ways in which AI is reshaping what journalism looks like and how audiences discover it.
Violaine Degasfrom Les Échos described a shift from an "article-centric" model to a more "service-centric" one, using AI to power translations and audio playlists. Le Parisien's Sophie Cassam Chenaïdiscussed using AI for video suggestions and insertions.
At La Provence, AI now produces audio versions of articles, summaries of top headlines, and personalised news digests - all voiced by synthetic voices cloned from actors.
Meanwhile, Romain Damery, Senior Director of Technical SEO at Amsive, discussed the ongoing evolution of search discovery. With AI Overviews in Google already impacting click-though rates, he said visibility will increasingly depend on AI-optimised content: clearly structured, uniquely insightful, and credibly attributed.
As publishers adapt to rapid technological shifts, their willingness to experiment, collaborate, and invest in people will be vital in shaping AI's role in journalism.
Or, as The Independent's Broughton put it, in these uncertain times, publishers would do well to "channel all our efforts into being the master of our own fate, by using our own terms."