Who controls the information channels? Media ownership, concentration, and the limits of influence

Ownership and concentration: measurable forms of power

If a company owns a newspaper, broadcaster, or online outlet, the owner has real power. It can: buy or sell the outlet, appoint the board, set long-term business strategy, approve or cut budgets, choose senior management, merge newsrooms, close programs or entire brands and change the structure of incentives and costs. Those are not trivial powers. But ownership does not automatically mean that the owner personally determines every headline, every question a journalist asks, or every sentence in a report. Between the owner and an individual article there may be a board, editor-in-chief, section editors, professional rules, contracts, unions, regulators, public scrutiny, competition, and the reputation of the brand.

We therefore need to distinguish: ownership → potential for influence from: documented intervention → evidence of specific influence. The European Media Freedom Act treats both dimensions separately for exactly this reason: it requires ownership transparency while also addressing the protection of editorial decisions from improper influence by owners, the state, and other interests.[3] A media market can contain many titles while still having relatively few real owners.

If one corporation owns several television channels, radio stations, production companies, websites, and distribution systems, the public may see many brands while the economic control belongs to a single group. The Media Pluralism Monitor therefore does not simply count outlets. It assesses: horizontal concentration within a sector, cross-media concentration, transparency of ownership, the influence of owners and commercial interests on editorial work, concentration of digital platforms and the financial sustainability of media organizations.[2][4]. MPM 2026 found that market concentration remains one of the most important structural vulnerabilities in the European media environment.[2] The important consequence is:

*a plurality of logos is not necessarily a plurality of owners, but a common owner is not automatically proof of a single editorial line either.* The second claim requires additional evidence.

An owner can influence a newsroom without calling the editor

Influence does not always take the form of an order saying “publish this story.” It can appear earlier in the process: which kind of journalism receives a budget, which foreign bureaus are closed, who becomes editor-in-chief, whether a months-long investigation is funded, how much legal risk the company is prepared to carry, whether the priority is local reporting, political commentary, entertainment, or short-form video, which audience segment the company wants to target and how dependent the newsroom is on advertisers or subscriptions.

This kind of influence is harder to establish because it often appears through a structure of decisions rather than one single document. That is why useful evidence includes: changes in editorial policy after an acquisition, documented interventions by an owner or board, personnel changes, internal instructions, budget changes, testimony from multiple independent employees and a pattern that extends beyond one disputed article. Ownership is a reason to examine conflicts of interest. It is not, by itself, the final proof of what that conflict produced.

The corporate map changes: consolidation and deconsolidation

The year 2026 provides a current example of how quickly the map of information channels can change. Paramount Skydance already includes CBS, CBS News, CBS Sports, Paramount, Nickelodeon, MTV, BET, Comedy Central, Showtime, Paramount+, and Pluto TV.[5] In February 2026 it signed an agreement to acquire Warner Bros. Discovery, whose portfolio includes CNN.[6] As of 24 September 2026, the transaction had not yet formally closed, although it had received numerous regulatory approvals and moved closer to completion following a settlement with U.S. states.[7][8] Part of the U.S. settlement also contemplated a specific structure for the editorial independence of CNN and CBS News.[7] This example does not prove political coordination of news coverage.

It proves something else that is measurable: *one corporate transaction can bring two of the most recognizable U.S. news systems under one controlling structure.* That is a legitimate media-plurality issue even without any theory of secret editorial orders. Concentration does not move in only one direction. In January 2026, Comcast completed the spin-off of VERSANT as an independent company. CNBC, MS NOW, USA Network, Golf Channel, E!, SYFY, Fandango, Rotten Tomatoes, and other brands moved into the new company.[9] This is a useful reminder: an ownership map should not be frozen in a single year.

A chart that was accurate in 2023 can already be wrong in 2026. A serious analysis should therefore always state: the date of the ownership snapshot, whether a transaction is signed or actually closed, who has voting control, who is only a financial investor and whether specific editorial safeguards exist.

Distribution power: platforms, algorithms, and visibility

The biggest change of the digital era is that a news organization often no longer controls the path from the article to the reader. The Reuters Institute found in 2026 that social and video networks were used for news by an average of 54% of respondents across 48 markets, compared with 51% for news organizations' own websites and apps. Facebook was used for news by 43% of respondents, YouTube by 34%, Instagram by 26%, and TikTok by 20%.[1]

That creates a new kind of power. A platform can affect: which links are visible, which video is recommended, which account is suspended, how much organic reach a publisher receives, which content is demonetized, what appears in search results, how content is summarized and whether a user needs to visit the original publisher at all. This is distribution power. It is not identical to editorial power, but for the reach of a story it can be just as important.

A traditional editor consciously chooses stories according to editorial criteria. An algorithm usually optimizes mathematical objectives such as: probability of a click, watch time, engagement, relevance, personalization, safety and moderation rules and the platform's business goals. The result, however, can function similarly to editorial selection: some content receives more visibility and some less. Ofcom's analysis of online news diets found an association between heavier reliance on online intermediaries — especially social media and search engines — and lower diversity of topics encountered by users.[10]

That is an association in the data, not proof that every algorithm is intentionally designed to narrow users' viewpoints. But it supports a key point: *the mode of distribution can change the information selection even when the article itself remains unchanged.*

The search engine as information gatekeeper

When users do not know which source they want, they often do not go to a news outlet; they go to a search engine. Pew found in March 2026 that, during a major breaking-news event, 28% of U.S. adults first turn to a search engine, 19% to social media, and 36% directly to a preferred news organization. In 2018, 54% went directly to a news organization and 15% to a search engine.[11]

That means part of the power has shifted from the editor to the gatekeeper of discovery. A U.S. federal court found in an antitrust case that Google unlawfully maintained a monopoly in general web search. In its 2025 remedies release, the Department of Justice stated that Google had for years handled roughly 90% of U.S. search queries.[12] That is an exceptionally strong market position. But it does not follow that every search result is politically ordered.

The correct evidentiary statement is: high concentration in search gives one system extraordinary influence over information discovery; a claim of intentional manipulation of a particular topic requires additional evidence about rules, interventions, or outcomes.

Advertising, monetization, and economic control of reach

A media organization can remain legally independent while being economically dependent on someone else's infrastructure. Digital advertising is an important example.

The UK Competition and Markets Authority had already found very high concentration around Google and Facebook/Meta in its major study of digital advertising.[13] In the United States, a federal court ruled in April 2025 that Google had illegally monopolized certain open-web advertising-technology markets; in September 2026, substantial remedies were ordered concerning interoperability, data sharing, and limits on self-preferencing.[14][15] This matters for journalism because advertising infrastructure helps determine: how much revenue a publisher receives, how much is taken by the intermediary chain, which users can be monetized, what data the advertising model requires and whether a small publisher can compete.

A platform can therefore influence the information system without writing a single story. If it changes the economics of distribution, it can indirectly change which media organizations survive at all. Consider two scenarios. In the first, a platform deletes an article. In the second, it does not delete the article, but: does not recommend it, excludes it from the advertising system, imposes additional conditions, reduces the reach of links, pushes it far down in search results and shows the user a summary so the original source is never visited.

In both cases the article technically exists. But its actual reach and economic value can be entirely different. That is why analysis of information power must measure more than formal censorship.

The fourth layer: operating systems, app stores, and infrastructure

The information flow has an even deeper layer. Before a user opens a news app, the user may already depend on: a mobile operating system, an app store, a browser, a search engine, a video or social network, advertising infrastructure, cloud services and a hosting or content-delivery provider. Under the Digital Markets Act, the European Union has designated major companies including Alphabet, Amazon, Apple, ByteDance, Meta, and Microsoft as gatekeepers for a range of core platform services.[16] The term gatekeeper matters.

It does not mean “owner of truth.” It means that a service is important enough as a gateway between businesses and users that its rules can systematically affect access to the market.

The state, public service, and private ownership

When discussing private corporations, it is easy to forget the other side: the state. A state can influence the media ecosystem through: legislation, licenses and frequencies, public-service media, subsidies, state advertising, taxation, access to official information, accreditation, concentration rules, sanctions, court proceedings and national-security restrictions. This does not mean that every publicly funded outlet is “state propaganda,” nor that every privately owned outlet is independent. It means that in every system we should ask:

who provides the money, who appoints the leadership, who sets the rules, and what safeguards are intended to prevent abuse? The EMFA introduced, among other things, transparency requirements on ownership, public funding, and state advertising, together with safeguards for editorial independence.[3] A public media organization does not have a conventional private owner, but it can become dependent on political appointments or funding. A private outlet has an owner, but its editorial structure can be well protected from direct intervention.

The ownership model alone therefore does not determine independence. We need to examine: how leadership is appointed, the length and security of mandates, the funding model, internal editorial statutes, mechanisms for dismissal, public disclosure of interests, the ownership structure and the actual historical record of interventions. European rules therefore address public-service-media independence and private-media ownership transparency as distinct issues.[3]

Slovenia: a small market, highly sensitive to concentration

Slovenia is a useful case because its market is small. In its 2026 Rule of Law Report, the European Commission wrote that Slovenia's media market remains concentrated and is particularly sensitive to concentration trends because of its small size and limited financial resources. Citing the 2026 Media Pluralism Monitor, it reports a very high risk for plurality of media providers and high risk for media viability and editorial independence from commercial or owner influence.[17] The new Slovenian media law introduced or strengthened: assessment of media concentrations, an ownership database covering direct, indirect, and beneficial ownership and transparency of state advertising and other forms of public funding.[17].

This does not mean that the Slovenian information space is “under one controller.” It means that the official European monitoring framework identifies structural concentration risk as a real problem that should be measured and supervised.

Platform concentration and the arrival of AI intermediaries

The modern paradox is this: there may be thousands of independent websites, podcasts, channels, and creators, yet a large share of their audience reach still runs through a small number of platforms. Ofcom found in 2025 that users in the United Kingdom spent roughly half of their online time on services owned by Alphabet and Meta.[18] The Reuters Institute reported in 2026 that news use is shifting toward social and video networks and that growth in online video is taking place mainly on third-party platforms, not on publishers' own websites.[1]

Two kinds of plurality can therefore coexist:

• high plurality of content producers;
• low plurality of distribution infrastructure.

That is one of the defining transformations of the twenty-first-century information system. In 2026, the Reuters Institute measured that about 10% of respondents across 48 markets used an AI chatbot for news, rising to 16% among people under 35.[1] That is still less than traditional social networks and search engines, but it introduces a new question: users can receive an answer without directly encountering the original source. An AI system can: choose which sources to use, combine several sources, omit details, summarize inaccurately, answer in the language and format the user wants and direct or fail to direct traffic to the original publisher.

This means that, alongside the editor and the social-network algorithm, a new synthetic information intermediary is emerging. For future pluralism, it is therefore no longer enough to ask only “who owns the newspaper?” We also need to ask: *who operates the system that tells a person what is in the newspapers?*

Does concentration automatically mean unified propaganda?

No. That would be too large a leap. A large corporation can own several media outlets that: have different editors, serve different audiences, compete with each other, operate under different professional cultures and publish mutually contradictory commentary. Likewise, two outlets without common ownership may publish similar stories because of: the same underlying event, the same news agency, similar journalistic routines, identical deadlines, the same public documents, similar economic incentives and competition for the same audience.

Therefore similar content ≠ proof of common command. But the opposite error is equally problematic: because there is no proof of a single command center, concentration does not matter. It does matter. It simply needs to be demonstrated at the correct level.

What would real evidence of coordinated editorial control look like?

If we want to claim that an owner, state, or platform coordinates content, we should look for more direct traces:

1. written editorial instructions;
2. emails or meeting records showing content intervention;
3. contracts or funding conditions tied to an editorial line;
4. documented punishment of journalists because of specific content;
5. systematic personnel replacement for political or business reasons;
6. technical documents showing intentional ranking changes for a particular topic or group;
7. communication between nominally independent actors about a shared content objective;
8. internal metrics and orders showing that the outcome was planned.

The stronger the claim, the stronger the bridge must be between capacity to influence and evidence of actual influence.

Seven different kinds of “control”

The word control is too broad. In practice, it should be separated into categories: 1. Ownership control Who holds capital, voting rights, and the power to appoint the board? 2. Editorial control Who decides what is published and how the story is framed? 3. Distribution control Who determines which content reaches users? 4. Monetization control Who determines which content generates revenue and under what conditions? 5. Infrastructure control Who provides the search engine, platform, app store, operating system, hosting, or cloud services? 6. Regulatory control

Who sets laws, licenses, sanctions, and market rules? 7. Cognitive influence How do algorithms, repetition, social signals, and source selection affect what receives attention? If all seven are collapsed into one word, the result is a dramatic but evidentially weak theory. If they are separated, the result is a map of actual power.

The largest shift: from controlling content to controlling attention

The classic model of media power was: owner → editor → content → audience. The modern model looks more like: source → editor → outlet → search engine/social platform/video network/AI → algorithm → advertising system → device → user. Something can change at every point. That means the most powerful actor is not necessarily the one that writes the story. It can be the actor that determines: what will be discovered, what will be recommended, what will be monetized, what will hold attention, which source will receive traffic and which business model will survive.

From this perspective, “control of information” often means control of distribution bottlenecks, not direct authorship of the content.

The THY-REALITY test and conclusion: who controls which gates?

The same procedure can be used for any important media organization or information channel. A. Ownership

• Who is the direct owner?
• Who is the ultimate or beneficial owner?
• Who has voting rights?

B. Governance

• Who appoints the board and editor-in-chief?
• What editorial safeguards exist?

C. Money

• Does revenue come from subscriptions, advertising, the state, donations, or a parent company?
• Who is the largest advertiser or funder?

D. Distribution

• How much traffic comes from Google, Facebook, YouTube, TikTok, or other intermediaries?
• Is the outlet commercially dependent on them?

E. Content

• Are there documented interventions?
• Does a change of ownership coincide with a measurable change in editorial policy?

F. Comparison

• Is the same fact confirmed by primary documents and competing sources?
• Is the story merely a rewrite of another outlet's report?

This procedure is more useful than asking: “Which media outlet is good and which one is bad?” The evidence does not support a simple picture in which a single owner or organization centrally manages the entire global information system. It does support something more concrete. Media ownership is concentrated. Digital distribution is even more concentrated. Search, social networks, video, digital advertising, app stores, and emerging AI systems create new information bottlenecks. Regulators in the EU, United Kingdom, and United States treat these bottlenecks as real questions of competition, pluralism, and editorial independence.[2][3][12][14][16]

The most accurate answer to “who controls the information channels?” is therefore: *different actors control different parts of the channel.* An owner can control capital and governance. An editor controls content. A platform controls discovery and reach. Advertising technology controls revenue. An operating system and app store control access. The state controls part of the regulatory framework and some funding. AI increasingly becomes another layer between the original source and the person.

The real power of the modern information system is not necessarily located in one secret room. It lies in the concentration of several bottlenecks through which an enormous amount of apparently plural content has to pass. And that is why serious research must always separate: who has the capacity to influence, who actually influenced, and what evidence supports each specific claim.