Dr Jan Erik Kermer is co-lead Researcher of STOA’s report on “A European news streaming platform. Study on European added value and governance”.
Why do we need a European news platform when people can already watch news from other countries online; are EU citizens interested in news from other EU countries?
Jan Erik Kermer: The platform addresses a discoverability problem rather than content scarcity, aiming to surface high-quality European news in a media landscape dominated by engagement-driven algorithms that prioritize clickbait over critical information.
Based on a review of comparative empirical studies, traditional media provides minimal European coverage, with roughly one in ten news stories focusing on the EU and even fewer on neighbouring countries. Supporting this, a survey (https://europa.eu/eurobarometer/surveys/detail/3153 ), conducted by the European Parliament, found that around 30% of citizens hadn’t read, seen or heard any EU-related news recently.
This platform would serve as a central focal point for high-quality content from neighboring countries in multiple EU languages, focusing on local and transnational news rather than Brussels-based coverage to potentially engage wider audiences beyond the policy community. However, whether this approach can successfully attract broader public interest remains an open question, as past European media projects have struggled with audience engagement outside specialized circles.
The study claims this platform could strengthen European democracy – how would watching news from other EU countries make me a better citizen?
Jan Erik Kermer: The platform would break citizens out of their “national “rabbit holes” by providing multiple perspectives on issues, such as allowing Spanish citizens to see how other countries report on Spanish affairs, which strengthens democratic discourse through diverse viewpoints. Further, it addresses the current fragmented online information ecosystem where search engines are designed to show mainly domestic news, preventing Europeans from understanding common challenges and developing solidarity across borders. A cross-lingual search function would allow users to find coverage on topics like climate change from news outlets across different countries, with algorithms designed to ensure smaller nations’ voices are heard alongside major players. This approach aims to transcend national borders, reduce biases, and create stronger awareness of shared European identity and common challenges.
How would you balance content from large countries like Germany and France with smaller nations like Malta or Estonia?
Jan Erik Kermer: The platform would ensure inclusive governance with representation from all member states to avoid domination by larger countries like Germany or France, making certain that smaller nations like Estonia and Latvia, for instance, have a voice. Algorithms would be designed to promote “serendipitous content” from countries users might not typically engage with, encouraging exposure to coverage from nations they haven’t previously clicked on. This algorithmic approach would be crucial for maintaining a pluralistic information environment where smaller countries’ perspectives are actively surfaced rather than overshadowed by content from major EU member states. The goal is to prevent the platform from replicating existing biases that favor larger member states while ensuring diverse European voices are heard.
How accurate would AI translations be and how would you prevent misinformation, especially when using automatic translation and little to no human oversight?
Jan Erik Kermer: AI translation accuracy remains a significant challenge, particularly for “low resource” languages with limited training data, which can result in lost nuance, context errors, and mistranslated acronyms or sensitive terms. Mitigation strategies include using local European large language models, clear labeling when AI technologies are being used (as required by the upcoming AI Act), and mechanisms for users to flag translation errors. Content providers could add context boxes for sensitive issues where mistranslation risks are higher; however, full human oversight is not feasible as it would be too labor-intensive for streamed content. Users must understand the platform cannot be their primary news source, as although it could offer similar reliability to other online translation programs, It cannot fully guarantee that information has been verified after translation.