Editor’s note: we are republishing a note that previously appeared in The Fix Media’s weekly newsletter. Subscribe to get everything you need to know about the European media market every Monday, along with monthly special reports.

Building a pan-European news site has been a notoriously difficult task. It’s not that no one tried, but existing players have mostly been either niche or not truly pan-European.

Three years ago David Tvrdon wrote for The Fix about “why the pan-European news site idea is so intriguing and the reality so impossible”.

The infeasibility comes down to several factors, chief among them the language barrier and cultural differences. “Instead of spending more time on the breadth and quality of the pieces, you need to put your energy into feeding each large part of Europe with its own language”, David groused.

One factor that has changed since we ran this column, though, is the rapid growth of generative AI. While AI isn’t capable of replacing the reporter, it’s increasingly augmenting, if not replacing, translators – and breaking language barriers before our eyes.

In announcing their partnership with OpenAI, Le Monde touted the launch of an English version that relies on DeepL (and human proofreading). In a new story for The Fix Priyal Shah cites the case of Yle, a Finnish broadcaster that has leveraged AI to translate its stories for Ukrainians who found refuge in Finland following Russia’s invasion.

Another possibility, a more hypothetical one for now, is personalisation. Imagine a pan-European news outlet that can dynamically adjust its content and user interface to each reader’s location, language preferences, interests, and behaviours.

For example, a Belgian entrepreneur interested in agriculture could receive a newsreel featuring EU policy updates, analyses of sustainable farming innovations from the Netherlands, and profiles of agtech startups from across the continent. Meanwhile, a Bulgarian programmer might see the latest on European tech regulations, funding news from the Baltics, and an interview with the winner of a Romanian hackathon.

This customisation could go beyond typical article recommendations to include optimising elements like reading level, article length, tone, and multimedia formats to match individual preferences.

To be sure, the technology isn’t fully there to enable this example. Neither is it good enough for large-scale AI-powered translation that requires little human presence. But it increasingly feels like the question of “when”, not “if”.

Source of the cover photo: Felicia Buitenwerf via Unsplash

translation

Latest in translation

Can AI help build a pan-European news publisher?