Monopolization in the AI interface and media in AI

Media publishers around the world share a common problem. Because the media industry, which lost control over distribution due to search engines and social media, is now seeing the ground beneath it slip away even more every day due to AI. As I have written before, the widespread adoption of tools like ChatGPT, Sora, and Meta AI means that the entire distribution system of the media must be rebuilt and reorganized.

Following the start of the collapse of Google's search engine model, the new search model it launched with AI and the new projects of other technology companies regarding AI searches and AI interfaces have sparked new debates. When this happens, publishers, who will no longer be autonomous structures capable of reaching their own users directly, will turn into content suppliers dependent on the interfaces provided by AI. To understand this transformation, it is necessary to look at it through the framework of the political economy of media. Below, I will examine this rupture in general.

CHANGING DYNAMICS OF THE AI ENVIRONMENT FOR PUBLISHERS

With the leap in AI usage, media consumption, especially among the younger generation, has undergone a behavioral change. Classic media sources are being replaced first by social media and now by AI interfaces. Chatbots are becoming the new interface for inquiries. Young people no longer search for information on the web; they obtain it through AI output. While every new technology environment transforms operations, news consumption is becoming mediated by AI. This means that whoever controls the interface will also control the flow of information and news. It is predicted that by 2028, AI search will surpass classic search engine searches. In a structure where news turns into data and content within the interface, the new gatekeepers in the publishing field will be the companies that produce AI and its interface. So, will the industry, which has relied on revenue generated from the search engine business model for many years, fall into the same trap?

RELIABILITY ISSUES OF AI ANSWERS

EBU research shows that approximately 45% of AI answers contain incorrect information (ebu.com). Source problems, hallucinations, and timeliness issues persist. This actually exacerbates the problem that Clay Shirky wanted to explain with the concept of "algorithmic authority" in the sociology of knowledge. AI does not disclose its source, determines the information it selects itself, and there is no public mechanism to audit its errors. Algorithms that serve as inquiry interfaces gain authority even when they are wrong. This is because the user sees the speed and fluency of the answer, not the sources behind it. The fact that the produced information is incorrect, is reproduced in hidden AI layers, and that AI systems feed their errors back through their own training cycles ultimately means the decay of information.

DEEPENING OF THE ''WINNER TAKES ALL'' RISK

In the political economy literature, platform capitalism (Srnicek) fundamentally describes how platformization and monopolization in the platform ecosystem are increasing rapidly. The establishment of browser, assistant, and application ecosystems by AI companies is the new stage of platform capitalism. Today, with large corporate media companies taking more space on AI platforms through licensing and copyright collaborations, the invisibility of small publishers, and the monopolization of content circulation and distribution while content production increases, further triggers the "winner takes all" situation in the digital economy. As in the case of the OpenAI–Schibsted agreement (schibsted.com), the articles and opinions of certain publishers are highlighted in AI interfaces. This throws the "neutral technology" fairy tale into the trash. In this context, producing the best content will not be enough; it will be necessary to have the best interface layer. The fact that some publishers' traffic has dropped by 40% due to Google's AI summaries is the first practical proof of this. The publishing sector does not have the luxury of continuing to view this transformation merely as a traffic decline. However, today, even many large global media companies do not have the financial infrastructure to build foundational AI models. The sector's choice to avoid falling behind is again the ready-made solutions of AI companies.

TRANSITION FROM AI IN MEDIA TO MEDIA IN AI?

The answer to this question is briefly related to the subjectification of the environment in media theory. It is that AI becomes the channel for accessing content, the shaper of content, the distributor of content, the personalizer of content, and the actor that exerts commercial control over content. In other words, users do not go to websites; websites come to the user via AI. With systems integrated into the browser, AI becomes a media consumption interface and even an operating system. While content becomes a component of the AI answer, the user receives content through personalized AI assistants. Thus, news publishers dissolve within AI, breaking away from the environment, context, and brand value. When this happens, media is inherent to the AI platform and is its content.

THE RISE OF AI AS AN ACTOR: THE AGE OF NEW INTERMEDIARIES

Autonomous actors are also acting units. According to actor-network theory (B. Latour), entities other than humans—in this case, AI and machines—are actors, and agency can be delegated between humans and technological objects. That is, a task, decision, or action can be loaded onto an object/technology. For example, AI as an actor is becoming capable of deciding on our behalf which information to look at, which content we will find "reliable," and which news we will never reach. The results in publishing are a decrease in human reader traffic and an increase in AI traffic. For the publisher, it is the shift of the true addressee of the content from human to AI. It is the further shift of the news value chain, which has already been transforming for a long time, to intermediary layers. In this case, publishers are forced to optimize their content and news for both human users and AI. This will shift the "existence of a debate only among the elite" finding in W. L. Bennett's indexing theory regarding decision-making in news among political elites and policymakers to the competition between global technology companies. As a result, media left to the mercy of companies producing AI technology will switch to a dual indexing system. That is, a dual indexing to be done according to the human end-user and the AI actor.

In short, the bitter truths facing publishers are these: Content is important, but the interface through which content is accessed is becoming more important. AI companies that control the AI interface will become the ones that control the news system. While the media loses ownership of the value it produces and delegates its gatekeeping and public sphere role to interface companies, it will have to optimize its content largely according to AI platforms and companies to maintain its economic position.

To summarize with scientific concepts:

• This is a concentration of platform capitalism.

• This is the construction of algorithmic authority.

• This is a process of digital information enclosure.

• This is the transfer of the field of communication, news, and information to a few global AI companies.

If the media focuses only and still on traffic loss during this process, it will have been a spectator to the liquidation process of its own existence. The process where information producers become invisible and invisible information processors—AI platforms and companies—control the news system through interface power can be read as the new and final stage of the commodification of information from a political economy perspective.

Those who look at this picture optimistically and say only "digital transformation is necessary in media" are wrong. This picture is the increase of digital monopolization within the media and even the digital domination of technology companies over the media industry and the publishing sector.

Let me finish with one sentence:

Whoever controls the interface controls the production, distribution, and revenue of information, and today, the interface belongs to AI companies.