News
AI in the Newsroom: UNESCO supports ethical integration in South-East Europe
Artificial Intelligence (AI) is revolutionising journalism, offering innovative tools for streamlining newsroom operations and expanding media reach. Experts warn of high risks, from misinformation to hate speech, censorship, and advanced disinformation tactics like deepfakes, among others.
To tackle these challenges head-on, UNESCO, Reporters Without Borders (RSF), and the Journalism Trust Initiative (JTI) brought together different voices on 6 May 2025, at the Press Club Brussels Europe, in Belgium. Marking World Press Freedom Day, with the theme of 鈥淩eporting in the Brave New World: The Impact of Artificial Intelligence on Press Freedom and the Media鈥, the round table explored Generative Artificial Intelligence鈥檚 (Gen-AI) integration into South-East Europe鈥檚 media sector through self-regulation.
Over 25 representatives from media outlets, universities, and press councils across the Western Balkans joined European Union institutions (DG ENEST, DG CONNECT), as well international organisations to discuss AI鈥檚 transformative impact on journalism in a region facing fragile media environments, government interference, and inadequate digital regulations.
Among the voices was Andris Kesteris, Principal Adviser for Civil Society and Media at European Commission鈥檚 Directorate-General for Enlargement and the Eastern Neighbourhood (DG ENEST). Kesteris argued that self-regulation could be more effective than formally introduced legislative norms by accompanying the development of AI in the realm of journalism rather than curbing it.
The EU Sustain Media programme, to be launched in September 2025, will assist in this process. The EU should provide financial support to help introduce AI into the daily operation of independent media in the Western Balkans.
Through the EU-funded project 鈥Building Trust in Media in South-East Europe: Support to Journalism as a Public Good鈥, UNESCO is actively promoting the ethical and transparent use of Gen-AI in both media and social media platforms in the Western Balkans. A milestone came last December at a regional conference where UNESCO supported the adoption of a landmark statement and recommendations on AI integration in newsrooms.
For Tawfiq Jelassi, Assistant Director-General for Communication and Information of UNESCO, this year鈥檚 World Press Freedom Day was more than symbolic; it was a unique opportunity to reflect together on AI鈥檚 impact on press freedom, media pluralism and the safety of journalists, women and men.
This is also a moment to reaffirm our engagement to promote human oversight, transparency and ethical standards of AI technologies in the digital space, through a multistakeholder approach.
鈥Media self-regulation is a critical tool to help navigate the challenges posed by AI,鈥 stated Borislav Vukojevi膰, UNESCO Consultant and Senior Assistant at the Faculty of Political Science (University of Banja Luka) reflecting on the opportunities and risks of AI, and tools for its ethical use.
One such framework is the Journalism Trust Initiative (JTI) certification, launched by RSF, which ensures that media outlets adhere to transparent, professional standards, in the use of AI in newsrooms. Similar efforts include the Serbian Press Council鈥檚 revision of its Code of Ethics to incorporate AI-specific guidelines, emphasising editorial accountability and transparency, and the clear labelling of AI-generated content.
A major challenge remains: raising public awareness of AI-related ethical breaches. In response, press and media councils in the Western Balkans and T眉rkiye have focused their annual awareness campaign on 鈥淢edia Ethics in the Time of AI鈥, producing 2 videos to engage the public.
Media representatives also advocated for a strategic approach to sustainable news revenue models, fair compensation for AI-utilised journalistic content, and early AI education for journalists and citizens. Finally, particular attention was given to the risk of biases within AI models, which can exacerbate gender stereotypes, digital sexism and broader systemic inequalities.
If AI is programmed or trained on data that is discriminatory towards women, minorities, or other marginalised groups, discrimination becomes systemic and automatic. We must understand that these tools鈥 responses cannot be accepted uncritically.
