INSURRECTION/ SP MPs rebel, drag Gjati to vote on territorial division!

 Kliko këtu për të tjera
Voxpress:

From code of ethics to algorithm

2026-09-18 12:53:54Pikëpamje SHKRUAR NGA ALIDA CENAJ
From code of ethics to algorithm
Alida Cenaj

How editorial responsibility is being transformed in the era of platforms and Artificial Intelligence

For more than a century, professional media has developed on a fundamental principle: the information that reaches the public passes through an identifiable editorial process. Behind it there is a chain – from the journalist and editor, to the editorial office and the media entity – that selects, verifies, shares and decides how the information is published. This responsibility has not only been legal, but also a trust agreement with the public: someone has chosen the news, assessed the sources, set the title and context and, if the standards are violated, can be held accountable. It is precisely this model that is being transformed. In the digital environment, the information experience is increasingly mediated by recommendation systems, ranking, profiling and algorithmic models that determine what becomes visible and what does not. Generative Artificial Intelligence adds a new dimension: technological systems can produce text, sound, images and video, which are then distributed through the same algorithmic architectures. Thus, a question arises that touches on the essence of media law and ethics: is the traditional concept of editorial responsibility still sufficient to guarantee transparency, pluralism, and public trust?

From information gatekeepers to attention managers
The history of the media can also be read as the history of control over the flow of information. In the traditional model, newsrooms exercised the function known as “gatekeeping”: selecting, filtering and prioritizing information before it reached the public. Technology has not necessarily changed the rules of news preparation in the newsroom, but it has significantly transformed the way it is distributed and visible. In other words, platforms do not decide what news a newsroom should produce, but they fundamentally influence which news appears to a wide audience, which is recommended and which is lost in the flow of content. From gatekeepers – information gatekeepers – we are moving towards an environment managed by attention keepers: systems that direct the public’s attention.

This distinction is fundamental. If traditional media competed primarily for audience and credibility, the platform economy competes for user time, engagement and data. Information is not always ranked according to public relevance, but also according to its ability to generate clicks, comments, shares, viewing time and user returns to the platform. Does this mean that the algorithm is an editor? In the legal and professional sense, no! It has no editorial responsibility, does not apply a code of ethics, does not search for news, does not verify sources and does not judge the public interest. However, algorithmic systems produce consequences similar to some editorial decisions: they rank news and videos, recommend content, increase or limit exposure and personalize the information experience.

As media and digital communication scholar Tarleton Gillespie notes, algorithms play an increasingly important role in selecting the information that is deemed relevant to us. Therefore, the traditional question “Who wrote or published this news?” is now supplemented by another: “Who decided that I should see this particular news?”

So the algorithm doesn't really replace the editor in producing and verifying information, but it can function as a kind of distribution editor. And in an environment where visibility determines impact, the decision on distribution is just as important as the one on publication.

The algorithm is not neutral
The phrase “the algorithm decided” gives the impression of a technical and objective process. But algorithms are designed by people and organizations, trained on data, and optimized for specific objectives, such as viewing time, interaction, or interest prediction. Therefore, mathematical coherence does not guarantee social, cultural, or democratic neutrality. Researcher Taina Bucher describes this through the concept of “programmed sociability”: online relationships are organized by technological infrastructure, and technological choices have consequences for the visibility, participation, and power of different voices. When a system favors content that provokes strong reactions, the goal may be to increase engagement, but the consequences can affect the tone of public debate and the way certain narratives are spread. While José van Dijck, Thomas Poell, and Martijn de Waal define platform society as an ecosystem “driven by algorithms and fueled by data.” Content passes through personalization and ranking systems that take into account the user's history, interactions, and predicted interests. Two citizens can open the same platform at the same time and face different information realities.

Personalization helps navigate the flow of information and gives small media outlets, independent journalists or local creators the opportunity to reach new audiences. But the same mechanism can fragment the information space: the public may no longer be exposed to the same facts and topics. The challenge here is not to oppose personalization, but to understand its criteria and its consequences for pluralism and democratic public debate. Two similar audiovisual contents on the same platform can have completely different reach – from a few hundred to millions of views. In linear media, the difference was mainly related to the editorial decision and the position in the programming; on platforms it is also influenced by the time of publication, interactions, viewing duration, audience and signals from the recommendation system. Editorial quality remains important, but it does not alone determine reach. Here the tension arises between editorial and algorithmic logic: the editorial office may prioritize public interest, while the algorithm prioritizes the content that holds the longest attention. Who, in this case, builds the real hierarchy of information?

From editorial responsibility to systemic responsibility
European legislation has not declared the editorial algorithm nor has it replaced editorial responsibility with technological responsibility. But the evolution from the Audiovisual Media Services Directive (AVMSD) to the Digital Services Act (DSA) and the European Media Freedom Act (EMFA) shows a broadening of focus: from content and the subject that publishes it, to the distribution architecture, moderation, recommendation systems and systemic risks. The AVMSD links media service to editorial responsibility, that is, to effective control over the selection and organization of programs. The 2018 review also expanded the scope to video distribution platforms, which do not exercise classic editorial responsibility over user content, but organize it through automated means. The DSA does not treat platforms as editors of all user content, but recognizes that their design and functioning can produce risks that cannot be explained by looking at a single post or video. EMFA places editorial independence and pluralism at the center. In the media-platform relationship, Article 18 provides procedural guarantees for media providers in the face of certain platform decisions that restrict or suspend their content.

So AVMSD, DSA and EMFA are an advanced response to the transformation of media, but not necessarily the final one. The law may require transparency on recommendation systems, but algorithmic complexity does not easily translate into understandable explanations for every citizen. Effectiveness depends on the quality of audits, access to data and the capacities of authorities. Transparency, too, is not the same as accountability: knowing that a system optimizes engagement is not enough without mechanisms to analyze and correct the consequences.

The challenge of professional media and the role of the regulator
Algorithmic transformation does not make the editorial office unnecessary, on the contrary, it places it in front of new responsibilities. Professional media must maintain standards even when content is adapted for platforms. The pressure for clicks, instant reactions or formats favored by recommendation systems should not replace the public interest. Knowing the logic of the platform is necessary, but it does not mean submission to it: a news story can be formatted without losing accuracy, a headline can be attractive without being misleading and a short video without taking the fact out of context. The challenge is to use technology, without allowing its criteria to become the media's code of ethics. This also applies to regulatory authorities, who must expand the way they understand the audiovisual environment. Monitoring content remains necessary, but it is not enough to understand how a material reaches the public and how it is amplified. The regulator should not control every algorithmic decision, but it can help build an environment where responsibilities are identifiable, complaint mechanisms work, user rights are respected, minors are protected and professional media maintain independence. Media education is also a long-term investment and today it should be expanded towards algorithmic education. In addition to traditional questions about the author, source, verification and the difference between news and opinion, the citizen should also ask why a content is being shown to him, what may have influenced the recommendation, which perspectives may be missing and whether the material is authentic or artificially generated. The goal is not for the citizen to become a programmer, but to understand that his information experience is mediated and that what appears on the screen does not necessarily represent the entirety or objective order of reality.

Times have changed, and so have information and responsibilities. Editorial responsibility has never been just about identifying the person who approves the publication of a news story. It links information and the power it represents with the obligation to be accountable. Technological transformation does not undo this principle, it simply makes it more complex. The algorithm is not an editor, but it affects functions that were once mainly performed by newsrooms: selection, ranking, repetition and visibility. The challenge is not to attribute to technology a responsibility that it cannot bear, but to not allow the organizations and people who design, use and control the systems to hide behind their complexity.

If the traditional media model built editorial responsibility around the question:

“Who published this content?”

The challenge of the platform age is to answer an equally fundamental question:

"Who decided that this content should reach me?"

The answer to this question will influence not only the evolution of editorial responsibility, but also how democratic societies will protect pluralism, freedom of expression, and public trust in the age of algorithms and artificial intelligence.

 


Në përputhje me standarded JTI

Video

SHIKO KANALIN NE YOUTUBE
×