From August 2, the rules set out in Article 50 of the European Artificial Intelligence Act (AI Act) impose new obligations on the transparency of content created or manipulated with AI. According to the European Commission, providers of artificial intelligence systems must enable the marking of generated or manipulated content so that it is distinguishable and detectable. Meanwhile, users of AI systems must notify the public when they encounter deepfakes or AI-generated textual content on matters of public interest, when this content is published without human review or editorial control.
The new rules aim to address one of the main challenges of the artificial intelligence era: the difficulty in distinguishing between authentic and synthetic content. The European Commission directly links transparency to the fight against fraud and manipulation, considering the identification of AI-generated content as an important element for preserving the integrity of the information ecosystem.
In this context, Hibrid.info's work and advocacy for transparency and responsibility in the use of artificial intelligence take on particular importance. Despite the fact that Kosovo is not a member of the European Union and is not directly involved in the implementation of these rules, adaptation to European standards and practices remains important for strengthening the integrity of information and preparing the country for the challenges posed by content created and manipulated with AI.
Over the past few months, Hibrid.info has consistently raised the issue of AI's impact on information integrity, arguing that the rapid development of technology must be accompanied by clear professional standards, verification mechanisms, and transparency towards the public.
In January of this year, Hibrid.info published the analysis "The information integrity crisis and the need to regulate the use of AI", which highlighted that the development of artificial intelligence technologies is making it increasingly difficult to distinguish between verified and manipulated content. According to the analysis, in the absence of clear editorial standards and control mechanisms, content created or manipulated with AI can circulate on social networks and then be republished as reliable information, increasing the risk of disinformation and public confusion.
This concern was also addressed through a roundtable discussion organized by Hibrid.info on January 30, 2026, where representatives of public institutions, media, academia and civil society discussed the impact of artificial intelligence on information integrity. The focus was on the need to build a common approach to the use of AI in the media and to address existing regulatory gaps in Kosovo.
A number of recommendations emerged from this roundtable, including encouraging voluntary labeling of AI-generated content, updating editorial and ethical codes, training journalists and editors, and building a national model of self-regulation and inter-institutional cooperation. Hibrid.info has presented this process as an effort to build ethical standards, guidance documents, and transparency models for the use of artificial intelligence in the media.
For Hibrid.info, the need for these standards is directly related to the media's responsibility to the public. In a time when photographs, videos, audio recordings and texts can be created or manipulated with few technological means, transparency is no longer just an ethical issue, but an essential element of trust in information.
Hibrid.info has also called for transparency in the use of AI in political communication, emphasizing that content generated with this technology is increasingly being used in election campaigns in Kosovo to illustrate electoral promises, visions for the future, and to comment on social and institutional developments.
The risk becomes particularly great when AI-generated content is used in political and social contexts. Synthetic content can be used to create false impressions, manipulate statements, present events that did not happen, or influence public perceptions of individuals and institutions. In these cases, the lack of labeling makes it even more difficult for citizens to critically evaluate the information they are provided with.
However, August 2 should not be seen only as a date when new rules begin to be implemented in the EU. For Kosovo, this date can serve as a reference point to start a more structured discussion on how AI-generated content should be managed.
Such an approach could start with immediate and actionable measures, such as voluntary labeling of AI-generated or manipulated content, updating codes of ethics and editorial guidelines, and training journalists to identify and verify synthetic content. At a later stage, these practices could serve as the basis for building a more sustainable regulatory framework.
Kosovo currently does not have a specific legal framework that establishes clear obligations for labeling content created with artificial intelligence in the media. For this reason, Hibrid.info has argued for the need for a coordinated approach, which would include the media, institutions, self-regulatory bodies, academia and civil society.