EUI’s Centre for Media Pluralism and Media Freedom joined the University of Dubrovnik in organising the 17th Dubrovnik Media Days and a Policy Lab on Information Integrity
 
The language of the food industry’s worst offerings – ‘junk food’, ‘pink slime’, and the like – has not only entered but transformed the way we critique media and information sphere, suggesting that what we consume mentally may be just as consequential as what we consume physically. In the same way that junk food is bad for our health, the junk in today’s news and information environment harms the health of democracy and individuals’ autonomy in forming their opinions and accessing accurate information. The 17th edition of the international scientific conference Dubrovnik Media Days – held from 22 to 24 October 2025 in Dubrovnik and co-organised by the University of Dubrovnik’s Faculty of Media and Public Relations and the EUI’s Centre for Media Pluralism and Media Freedom – focused on how generative AI is reshaping the integrity, economy, and ethics of the information space.
 
More than 40 scholars from a wide range of disciplines and three continents examined the structural implications of generative AI for journalism, democratic communication, and education: from its transformative potential to enhance transparency and accessibility of information to the risks of latent persuasion, bias, disinformation, and the economic disruption of media institutions as the central organisational and financial infrastructure for journalism. 
 
In today’s complex information environment, many things pass or are presenting themselves as media or journalism. This reflects a central paradox: journalism must remain open and inclusive in line with global standards on freedom of expression, yet it must also safeguard accuracy and accountability, the very foundations of its credibility, legitimacy, and democratic function. Several profound, technology-driven transformations have reshaped the very concept and economic foundations of journalism, its relationship with audiences, especially younger generations, and its interactions with new actors in the information sphere. Online platforms and influential content creators (such as podcasters and influencers) often operate in the realm of opinion sharing and opinion shaping, without investing in costly, time-intensive reporting. Nevertheless, they frequently enjoy greater reach and impact than traditional news media.
 
How can the integrity of the information sphere be safeguarded in an environment where:

• commercial media strategies frequently prioritise short-term click gains over journalistic quality, undermining long-term credibility and societal legitimacy;
• technology companies design and govern the core infrastructure of the information environment with limited accountability, often reinforcing fragmentation and polarisation – conditions fundamentally at odds with democratic needs;
• those same companies extract economic value from journalism without fair remuneration or even acknowledgment, weakening the sustainability of professional news production; and
• a growing ecosystem of news and political influencers operates without accountability, transparency, or professional standards, while often achieving greater reach and impact than established media?

 
These were the questions addressed at the conference with further spotlight on the challenges intensified by the rapid adoption of generative AI, which, if not responsibly governed with full respect for the fundamental right to freedom of expression, risks amplifying both economic pressures on newsrooms and the spread of low-quality or manipulative content. 
 
The CMPF team presented their work on infrastructure, economics, practice and regulation of news and information integrity:
 

• Jan Erik Kermer and Tijana Blagojev explored the rise of different types of “pseudo-media” websites in Europe – that is, outlets that mimic real news sources to mislead readers, usually for commercial gain or to manipulate public opinion. Their research was inspired by the lesser-known but growing phenomenon of “pink slime” websites – a US label borrowed from the processed, low-quality beef filler – which pose as local news outlets, churning out low-quality (typically AI-generated) content, often driven by commercial incentives (ad revenue) or serving political or ideological agendas. Beyond pink slime, they identified other pseudo-media types in Europe such as “doppelgänger” sites, state-aligned propaganda portals, “zombie” newspapers, “phantom media” and municipal media – which employ similar tactics to pink slime with the aim of misleading audiences. The authors argue that it is important to track how these insidious actors are infiltrating and undermining Europe’s fragile media landscapes. Their proliferation raises urgent questions about the health of Europe’s information landscape and how citizens can distinguish genuinely independent journalism from imitators. The issue is particularly pressing at the local level, where news deserts – that is, areas lacking sufficient, reliable and diverse information from independent local, regional and community media – make communities especially vulnerable to pseudo-media actors. 

• Iva Nenadić addressed the (infra)structural needs of the democratic information space and examined how generative AI, through its personalised applications, creates an “audience of one”, a model that conflicts with democracy’s requirement for a shared information space and a common reality. She presented this shift in her paper titled “Democracy Dies in Fragmentation: Generative AI and Hyper-Fragmented Audiences”, co-authored with Ivana Hladilo from the University of Dubrovnik’s Faculty of Media and Public Relations, and drawing on analytical work conducted within the Council of Europe’s Committee of Experts on the implications of Generative AI for Freedom of Expression.

• Roberta Carlini highlighted the key elements of the “new copyright wars”, after the mass development of GenAI systems, presenting a working paper written with Anya Schriffin (Columbia University) and Natalia Menendez (EUI). As recognised in the European Commission’s Action Plan to Support Recovery and Transformation – Europe’s Media in the Digital Decade, “the media sector as a whole covers a variety of businesses that produce and distribute content, that share synergies, and whose value is based on intellectual property”. Accordingly, the protection of intellectual property and the fair remuneration for its use and dissemination is at the core of the media sector’s sustainability and integrity. However, this legal protection has relevant shortcomings in the era of GenAI, when media content – as well as others creators’ content – has been freely used to train LLM, and the AI assistants’ answers tend to replace the direct access to the media content, foreshadowing the risk of a “zero-click future”.  Both US and EU legal systems fall short in addressing these new challenges. In the uncertainty of the current situation, many media companies negotiated with AI companies to license their content. But key questions remain: Is the price of these deals fair? What is their impact on media pluralism? And, is copyright protection itself the right tool to address the multiple challenges and risks journalism faces in the AI era?

• Elda Brogi presented the legal basis, context evolution, and evidence building that led to the adoption of the European Media Freedom Act that is now in full implementation across the EU but awaits specific national interventions that will ultimately decide its effectiveness. In order to inform and monitor this process, the Centre for Media Pluralism and Media Freedom has been running an Observatory on the European Media Freedom Act and regularly organises and promotes multistakeholder discussions on the interpretation and implementation of the Act (EMFA Talks). Following the publication of “The European Media Freedom Act: media freedom, freedom of expression and pluralism,” produced for the EP LIBE in 2023, CMPF is now working on a study for EP-CULT on the implementation of EMFA in EU member states.

 
The 17th Dubrovnik Media Days addressed the above issues not only through scholarly discussion but also in practice: the scientific conference was complemented by a Policy Lab on information integrity, fostering dialogue between researchers, policymakers, and practitioners on concrete strategies to safeguard the quality and credibility of the information ecosystem, including through implementation of the recent European Union regulatory toolbox, namely the Digital Services Act, the European Media Freedom Act, and the AI Act.
 
Selected papers from the conference will be published in a peer-reviewed special issue of the journal Media Studies.
The post Generative AI is reshaping the integrity, economy, and democraticness of the information space appeared first on Centre for Media Pluralism and Media Freedom.