The WAN-IFRA Paris AI Forum brought together key industry players to discuss the rapidly evolving role of artificial intelligence in the news ecosystem.
While last year, many publishers were still very skeptical and hesitant, this year it seems many publishers go “all in” and are embracing this disrupting new technology. Here are the major insights that could help shape your strategy for the future.
AI alone is not a strategy, but a part of a larger one
A recurring theme from the forum was that AI is not an isolated initiative but rather fully embedded into the core strategies of leading news organizations. As Pia Rehnquist, Editorial Director from Bonnier News Local, put it, “AI is not a strategy. It is integrated into our strategy.”
With over 150 newspapers, Bonnier has leveraged AI to improve operational efficiency, boost product offerings, and explore new, innovative news formats. What Bonnier News Local’s integration of AI shows is that AI integration is a marathon, not a sprint. Their print automation project, which resulted in a 25-30% cost reduction in its first year, took two years to fully integrate across 30 newspapers. This deliberate approach ensures a sustained, beneficial impact on business outcomes.
AI can help boost three main pillars
AI has found its application in the news industry under three main pillars: efficiency, revenue generation, and innovation.
Efficiency gains were most evident in processes like print automation and customer service. For instance, AI chatbots at Bonnier resulted in a 50% reduction in email queries. For decision-makers in technology and operations, this shows how targeted AI deployment can streamline internal workflows and significantly reduce costs.
In the area of revenue generation, companies like KSTA Medien, Ringier, Mediahuis, and Bonnier are leveraging AI to personalize content and support editorial decisions. Amélie, an AI editorial assistant integrated into Bonnier’s CMS, assists journalists with headline suggestions and content summaries, providing support where it’s most needed. This integration has enhanced the quality of content while allowing journalists to focus on creating meaningful stories that resonate with audiences.
Innovations in AI are also changing how news is consumed and interacted with. Bonnier’s experimentation with synthetic voices and conversational interfaces allows users to interact with articles, providing new ways to deliver traditional journalism in more accessible forms. Such innovations are crucial for product and marketing teams looking to engage younger audiences who value interactivity and personalized experiences.
AI brings opportunities that must be balanced with ethical considerations
While the opportunities are tremendous, the forum also emphasized the importance of ethical guidelines and caution.
As Rien van Beemen, CEO of Mediahuis Netherlands, highlighted, experimentation from within the organization is vital, but so are clear guidelines to ensure that journalistic integrity is preserved. Mediahuis, with a 21% profitability on news media, leverages this financial stability to explore AI-driven opportunities while carefully balancing the core mission of providing unbiased, high-quality journalism.
The overarching message was that AI will be a significant part of the future, but it must be guided by human oversight. AI is most effective when it augments human effort rather than replaces it. Whether it’s helping journalists craft better stories, providing personalized reader experiences, or even experimenting with AI-driven news chats, human expertise is still central to ensuring these innovations benefit the news ecosystem sustainably.
AI will not kill jobs, but we will need to adapt
Petra Ehman, Chief Innovation and AI Officer of Ringier showed an interesting analysis of the percentage of the population being actively employed in the UK since 1850. Through various industrial disruptions, the employment rate has been remarkably constant and may continue to be so even with the introduction of AI.
That said, we will certainly need to adapt and build new skills. During a stunning master class on new video generation capabilities, Stanislas de Livonnière, Head of the Data & Innovation team at Le Parisien, showed us examples of just how these new skills will be put to use. Indeed, he illustrated how with little input a very insightful short video can be created using tools like Midjourney, Heygen, Runway, Kling, and Elevenlabs.
AI is changing media’s future revenue model
The debate over who pays for news isn’t new, but the rise of AI has intensified the discussion. A critical issue now is how publishers can be fairly compensated for the content utilized by AI models.
In recent years, numerous prominent publishers have formed partnerships with AI firms like OpenAI and Perplexity, trading content access for revenue. However, as highlighted in Ezra Eamon’s recent presentation, such revenue-sharing agreements typically benefit only major publishers. This raises concerns for regional and niche publications, which often find themselves in a “take it or leave it” situation.
Several initiatives are emerging to address these disparities. For instance, Annelies Jansen, Chief Strategy Officer at ProRata, shared her company’s strategy. Endorsed by leading publishers such as The FT, The Guardian, and The Times, ProRata aims to extend equitable compensation to mid-sized publishers as well. Their model pays publishers each time AI utilizes their content, whether in style, wording, or ideas.
Similarly, Miso AI is pioneering an innovative approach by developing specialized models—for example, a travel guide AI based solely on a single publisher’s archive. As their Co-founder Lucky Gunasekara shared, they also support a pay-per-answer model, which moves beyond the traditional reliance on website click rates for publisher compensation.
These evolving models represent a move toward leveling the playing field, offering new compensation frameworks that could benefit a broader range of publishers.
The value of an AI mentor in newsrooms
To help journalists harness digital tools, NOZ/mh:n MEDIEN established an internal coaching team. However, they quickly realized the team was too small to support the numerous journalists needing frequent guidance and training effectively. In response, the team developed an AI-driven coaching bot capable of providing 24/7 support. This bot enables journalists to seek on-demand assistance, ask knowledge-related questions, and receive timely feedback.
Recognizing the cost and speed constraints of processing large data sets with AI, the team implemented a ‘retriever’ system. This system pre-filters and reduces the data before it is analyzed by the AI. This approach helps manage costs and improve processing speeds by ensuring only pertinent information is considered in AI analyses.
With the groundwork laid by the retriever system, the AI can operate more efficiently. The coaching bot uses this refined data along with provided instructions and chat history to generate responses and support for journalists. This not only makes the bot more responsive but also more contextually aware, enhancing its ability to assist effectively.
“What can AI not do?”
Over the past two years, newsrooms have embraced AI for tasks like automation, synthesis, and editorial support. But what’s next?
Pia Rehnquist, Editorial Director from Bonnier News Local, ended her session with a question: “What can AI not do?”. That was probably the most important question of the day as there are so many opportunities now possible with this new technology.
David Caswell, Founder of StoryFlows, envisions a future where journalism could become entirely AI-driven within five years. While the specifics remain uncertain, one thing is clear: the pace of change is unprecedented. AI adoption is outpacing the uptake of both the internet and personal computers, with 40% of Americans aged 18–64 already using AI tools. This rapid growth suggests that innovations will continue to accelerate.
Even amidst this uncertainty, certain guiding principles are emerging. As highlighted across several presentations at the forum, the most successful AI initiatives often come from cross-functional teams, prioritize user and audience feedback, and embrace continuous experimentation.
The future of journalism may be hard to predict, but these principles will likely shape its evolution.
Other Blog Posts
Stay on top of the game
Join our community of industry leaders. Get insights, best practices, case studies, and access to our events.
"(Required)" indicates required fields