Part I — Situation overview

On the afternoon of 18 August 2026 Duna Médiaszolgáltató Zrt. — the successor with effect from 26 July 2026 of the Media Service Support and Asset Management Fund (MTVA) — published on its website the complete, 482-page documentary material collecting the documented cases of the news-editing practice of the Hungarian News Agency (MTI). Foreign correspondents and news editors calling themselves the Pro Domo working group collected the correspondence, e-mails and exchanges of messages over more than ten years: the first document is from 2015, the last one from the day before this year’s election. The documentary material arranges 266 cases into ten categories — news items suppressed or made on instruction, materials thrown out, manipulated content, censored and false news. According to the publisher “the documents record as direct evidence the instructions of managers, as well as those news items whose publication was prohibited or whose content was modified”, and no professional objection was raised against these news items. According to the announcement authenticity was verified, and the redaction of personal data was decided by establishing the levels of responsibility. There is uncertainty around the extent: the publisher and Telex mention 482 pages, 24.hu 483 pages; the analysis below uses the publisher’s own figure.

The background to the case reaches back to the beginning of August, when the working group handed the documents over to the interim leadership appointed in July, which then published a 15-page selection and ordered an internal examination. Our earlier analysis, 482 pages and one comment, was then about the procedural deficiency of the personnel decisions: about the withdrawal within half a day of the newly announced management appointments. The present step is of a different nature. It is not a personnel matter but a question of records management and publicity: the whole of the evidence base has come out into the public, and now the stake is what will happen to it. Two further threads run in parallel with this. The interim leadership of the public media is examining the lawfulness of the payments to the departing chief executives: according to the answer given to a public-interest data request, the sum of the benefits connected with employment relationships terminated between the election of 12 April 2026 and 2 August was more than 201 million forints, of which Dániel Papp received a net 28.1 million and Anita Altorjai a net 7.2 million forints. And on 20 August — on the feast of the foundation of the state — the news programme restarts on M1, according to the publisher’s announcement in an “independent and credible” form.

MIAK’s reading: the disclosure is legitimate and professionally grounded, but the documentary material now made public guarantees nothing in itself. A 482-page PDF file is not accountability but raw material. The document proves that the day-to-day influencing of content worked by a procedural route, in the form of informal requests and verbal instructions — that is, precisely on the channel that no change of personnel closes. The timing of the publication — two days before the restart of the news programme — also carries the risk that accountability becomes a communications event: if openness is a gesture and not a rule, then the new public media steps into the same credibility trap it wants to come out of.

Part II — Foundations in the literature

The framework of principle for the topic comes together at three authors. The main proposition of Spin Dictators by Sergei Guriev and Daniel Treisman (economist at Sciences Po and political scientist at the University of California respectively, researchers of information-based authoritarian rule) is that modern technique of power builds not on open repression but on shaping the information environment — so the operation of the system is hard to prove precisely because it lives not in commands but in decisions that look like editorial judgment; the documentary material now made public is therefore of exceptional value, because it records the requests themselves. Marshall McLuhan (Canadian media theorist, who formulated the proposition “the medium is the message”) argues in his 1964 book Understanding Media that the social effect of a medium of transmission is given not by its content but by its structure — from which it follows that the source monopoly of a news agency in itself determines what can reach the reader, irrespective of who sits in the editor’s chair. Shoshana Zuboff (professor emerita of Harvard Business School, researcher of the power structure of the digital economy) speaks in her work The Age of Surveillance Capitalism of the “division of learning”: in the information society the decisive question is not whether knowledge exists, but who holds it, who decides about access, and who decides who decides — this is the concept that makes the accessibility of the 482 pages a policy question. The detailed treatment of the literature — by author, with quotations — is to be found in the section 6.4 Literature in detail.

Part III — MIAK’s concrete proposal

MIAK proposes three measurable measures which together make the disclosure into institutional change.

3.1 A searchable, machine-readable database out of the documentary material (within 60 days)

In its present form the 482-page file cannot be analysed: it is not searchable in its structure, it cannot be filtered by period, category or level of decision, and it cannot be aggregated. MIAK proposes that Duna Médiaszolgáltató Zrt. transform the 266 cases within 60 days into a uniform, machine-readable data structure — with fields for case identifier, date, category, news topic concerned, type of intervention and level of decision — and make it available as open data, maintaining the present order of redaction of personal data. From the point of view of the D2 open data programme and the D8 disinformation-detecting system the material is of double value: it is at once a document of accountability and training data for a pattern-recognition instrument that supports the identification of future patterns of influence. Let access be public and without registration, alongside the structured version with the original page images — in the logic of the “division of learning” (see 6.4.3) the regulation of access is itself the question of power, so it cannot be left to organisational discretion.

3.2 A written editorial statute excluding verbal instruction on content (by the next amendment of the media act)

The most important structural lesson of the documentary material is not that there was influencing, but that it was without a trace. MIAK proposes that the editorial statute concerning the news provision of the public media — within the framework of the next amendment of the media act, with statutory authorisation — lay down: an instruction on content may be given exclusively in writing, with an identifiable sender and registered filing; carrying out a request on content arriving on a verbal or informal channel is a breach of the rules; the editor may report every such case to a complaints body outside the newsroom operating under a public procedural order, without adverse legal consequence. This proposal applies the logic of the A5 whistleblowing system to the newsroom environment, and it is the direct implementation of the A7 media pluralism as an institutional guarantee programme point. An important delimitation of competences: the content of the statute is established not by a government decree but, on the basis of the authorisation of an act adopted by Parliament, by the internal rules of the public-service media provider — government communication and public-service news provision are two different functions, and the present documentary material is precisely about the running together of these two.

3.3 An annual, public measurement of media pluralism that binds the present government as well (first edition in the first half of 2027)

The third proposal is the least spectacular and the most important. MIAK proposes that a public, numerical survey of the state of the Hungarian media structure be prepared every year — under the KU2 media-pluralism monitoring programme point, building on the indicator structure of the Media Pluralism Monitor of the European University Institute, with domestic adaptation. Let the subject of the measurement be market concentration, the distribution of state advertising expenditure, the governance independence of the public-service broadcaster and the institutional protection of editorial autonomy. Let the publication be an obligation laid down in legislation, not a question of government decision — this is the only way for the yardstick to be valid for the next governments, including the present one. The KU7 disinformation resilience and media-literacy programme is the natural counterpart of this: the data of the measurement can also be used as public-education media-literacy material, because the cases now made public give concrete, documented examples of how distortion arises.

The three proposals are bound together by one principle: each of them creates the trace where there was none before. The database makes the trace of past intervention analysable, the statute makes the future one compulsory, and the measurement of pluralism measures the structural state. The framework of the literature supports precisely this: if the technique of power lives in decisions that look like editorial judgment (see 6.4.1), then the protection cannot be personal either, only procedural.

Part IV — Expected effects and risks

Dimension Expected effect Risk
The public The documentary material becomes a verifiable, quantifiable factual basis; both media-historical research and the clarification of responsibility can rely on it The categorisation decisions taken during structuring themselves carry an interpretation; if the categories are determined by the publisher alone, the database may distort
Newsroom The obligation of written instruction gives protection in daily work, and makes an attempt at intervention costly Being in writing may turn into formal compliance: the substantive request is pushed onto an informal channel, and on paper everything will be in order
Public administration The delimitation of government communication and public-service news provision receives a regulated form The delimitation entails a political cost for every government, so postponing implementation is the most likely form of failure
Clarification of responsibility On the basis of the documented cases the examination of employment-law and — in the event of a report, according to the decision of the investigating authority — criminal-law responsibility is better grounded If the clarification of responsibility stops at the level of subordinates, the systemic fault remains untouched

The main dilemma stands between speed and credibility. A rapid disclosure with communications value gives a political return, but it worsens the legal and institutional return: errors of redaction, misleading categorisation, concentration on individuals instead of on the systemic fault. The slower, regulated route — statute, complaints body, annual measurement — has a political return close to zero, but a lasting institutional return. The proposal tips over to the risk side if only the first of the three elements is realised: a public database without a regulated editorial procedural order will be just as vulnerable in the next cycle as in the previous one. The documentary material is itself the best evidence of this: the legal framework — Act CLXXXV of 2010 on media services and mass communication, the Public Service Code prepared on its basis and MTI’s Editorial Handbook — existed at the time of the documented cases. It was not the rule that was missing but the trace of the rule’s enforcement.

Part V — Measurability and summary

5.1 What is worth following? (proposed KPIs)

MIAK proposes the following performance indicators (KPIs, Key Performance Indicators) for following — these are proposals, not government undertakings:

  • Readiness of the database: the proportion of the 266 documented cases available in structured, machine-readable form — proposed target: 100 per cent by the end of the fourth quarter of 2026.
  • Rate of being in writing: the proportion of instructions on content in the news provision of the public media that are filed and come from an identifiable sender, relative to all recorded interventions in content — proposed target: 100 per cent in the first full year following the entry into force of the statute.
  • Operation of the complaints route: the number of reports arriving at the complaints body outside the newsroom and the proportion of cases closed within 30 days. A low number of reports is not in itself a good sign: it may also indicate that the channel is not working.
  • Existence of the measurement of pluralism: whether the first public, indicator-level Hungarian media pluralism report appeared by the end of the first half of 2027, and whether its methodology is public.

5.2 Summary

MIAK’s request in a single sentence: let the 482 pages not be an end point but an input. Of the interim leadership of the public media MIAK asks that it transform the documentary material made public into a searchable database within 60 days; of Parliament, that in the next amendment of the media act it give statutory authorisation for an editorial statute excluding verbal instruction on content and for a complaints route outside the newsroom; and of the public, that it read the disclosure not as a result but as a starting point.

Two MIAK foundational values move together here, and it is no accident that it is precisely these two. Transparency because the present situation is in the strict sense the test of transparency: the document has become public, but the knowledge inside it is not yet accessible — publicity and accessibility are not the same, and the difference lies precisely where policy work begins. And accountability because the mechanism now uncovered was the product not of an absence of rules but of an absence of enforcement: the code existed, the trace did not. A system in which a rule is kept alive only by voluntary compliance starts again from zero at every change of government — and from MIAK’s point of view this is the real lesson, irrespective of who is in power.

Part VI — Reasoning and further sources

6.1 The press framing by spectrum

The narrative choices of the source bands were sharply distinct on this day — not in the facts, but in whom the story is about.

The left-liberal band carried the content of the documentary material and emphasised its character as evidence. Telex concentrated on the internal structure of the document and on the concrete cases — among them that domestic political news does not appear in it, and that the names of Hungarian political actors, with the exception of Péter Szijjártó, do not appear — HVG listed the case categories and the examples, while 444.hu put the publisher’s announcement and the reactions within the profession at the centre. The framing in this band: this is evidence, not opinion.

The general public-affairs band divided its attention in two directions. 24.hu carried the documentary material and the restart of the news programme on 20 August in parallel, that is, it framed the story as an institutional transition: what was, and what is coming in its place. ATV highlighted the thread of public consultation. This framing demotes the documentary material to a background to a future event — a substantive choice, because it places precisely the question of future guarantees at the centre.

The economic band carried the financial dimension. Portfolio put into its headline not the documentary material but the payments to the departing chief executives and the internal examination: a payment of 28 million forints, a sum of more than 201 million forints, an examination of lawfulness. This is the narrowest but the most concrete frame — and the only one that works with immediately verifiable figures.

The pro-government and conservative band turned the story onto the method of publication. On the platforms of Mandiner and Magyar Nemzet the communications director of Fidesz, Bertalan Havasi, called the process “internal denunciation-mongering”, and shifted the emphasis onto the utterances of the present prime minister in newsroom matters. This framing does not dispute the substantive assertions of the documentary material but makes the entitlement to collect it and the procedure of the present leadership the question. From MIAK’s point of view this argument is not negligible: the yardstick of editorial independence indeed cannot be divided in time — if an utterance by a head of government was a problem earlier, it is one now too. This is precisely why the sender-independent obligation of being in writing entered proposal 3.2: the rule does not ask who intervened, but whether there is a trace of it.

6.2 Facts and data

Datum Value Source
Extent of the documentary material 482 pages (24.hu mentions 483 pages) Announcement of Duna Médiaszolgáltató Zrt., 18 August 2026
Number of documented cases 266 Duna Médiaszolgáltató Zrt., 18 August 2026
Number of case categories 10 Duna Médiaszolgáltató Zrt., 18 August 2026
Time range of the collection 2015 – the day before this year’s election Duna Médiaszolgáltató Zrt., 18 August 2026
Previously published selection 15 pages MTVA, beginning of August 2026
Payments connected with terminated employment relationships (12 April 2026 – 2 August 2026) more than 201 million forints Answer of Duna Médiaszolgáltató Nonprofit Zrt. to a public-interest data request (Media1)
Dániel Papp’s departure payment net 28.1 million forints the same data request, Telex, 18 August 2026
Anita Altorjai’s departure payment net 7.2 million forints the same data request, Telex, 18 August 2026
Hungary’s governance quality indicator (control of corruption) −0.17 World Bank Worldwide Governance Indicators, 2024

The legal framework, to which the working group itself also refers: Act CLXXXV of 2010 on media services and mass communication, the Public Service Code prepared on its basis, MTI’s Editorial Handbook and the Fundamental Law. This is in itself the strongest argument for proposal 3.2: at the time of the documented cases all of these were in force.

6.3 Policy dimensions

  • Culture (programme points and background material) — the governance model of public-service broadcasting and the institutionalisation of the measurement of media pluralism (programme point ID: KU2, KU7); the background material of the area treats the funding and governance models of public-service media — the British public-service charter system, the German public-law model — all of which are based on a written statute.
  • Transparency and anti-corruption policy (programme points) — media pluralism as an institutional guarantee (programme point ID: A7), the protection of newsroom whistleblowers (programme point ID: A5), and the measurement of the pattern of the technique of power (programme point ID: A9, A13).
  • Digitalisation and AI regulation (programme points) — the open data standard and the requirement of machine readability for making use of the documentary material (programme point ID: D2), and the use of the documented cases as a sample database (programme point ID: D8).

6.4 Literature in detail

6.4.1 Guriev and Treisman: Spin Dictators

The basic proposition of the volume is that the dictatorships of the twentieth century built on fear were replaced in the twenty-first century by the exercise of power built on the management of information: instead of open coercion, power maintains itself by shaping the structure of the public sphere. The essence of this technique is unprovability — the intervention does not appear in the form of a command but looks like editorial judgment, and therefore it is always defensible afterwards as a professional decision. The Hungarian documentary material is an exceptional source precisely because it records this very layer: not the news items published, but the requests and traces of decision out of which the published news item came. The practical consequence of the authors’ argument for the present situation is that removing those who intervened is not enough to dismantle the system — the channel of intervention has to be abolished, because the channel is available to the next actor as well.

📖 Source: Sergei Guriev – Daniel Treisman: Spin Dictators

6.4.2 Marshall McLuhan: Understanding Media

According to McLuhan’s best-known proposition the social effect of a medium of transmission is determined not by the content flowing through it but by the structure of the medium:

“the medium is the message […] because it is the medium that shapes and controls the scale and form of human association and action.”

McLuhan adds that the “content” of any medium all too characteristically blinds us to the character of the medium itself — and this observation is the most precise description of the present debate. As long as public attention is directed at the individual cases, at the concrete news items thrown out and at the names, the structural question — that a central news agency enjoys a source monopoly towards the smaller newsrooms, so that a decision taken at a single point takes effect on many channels — remains invisible. In the case of the MTI documentary material this means that the change of personnel and the clarification of responsibility are necessary but do not abolish the structural vulnerability: that follows from the persistence of the source monopoly, and can be handled only with procedural guarantees — with a written statute, with a complaints route leading outwards.

📖 Source: Marshall McLuhan: Understanding Media

6.4.3 Shoshana Zuboff: The Age of Surveillance Capitalism

Zuboff calls the “division of learning” the ordering principle of the information society, and breaks it down into three questions: who knows, who decides who may know, and who decides who decides. According to her argument, in the information age power derives not primarily from the possession of knowledge but from disposition over access — from the fact that somebody unilaterally determines to what depth a body of data is available to whom. Applied to the 482-page documentary material this framework sharply illuminates a difference that is blurred in the present debate: the document is public, but not accessible. In the case of an unstructured, non-searchable file the publisher remains the only actor who can survey and aggregate the material as a whole. This is why it is not a formal but a question of power whether the machine-readable data structure under proposal 3.1 is prepared: without it, the disclosure does not hand over disposition over access, it only apparently shares it.

📖 Source: Shoshana Zuboff: The Age of Surveillance Capitalism

6.5 International comparison

Operative realisations of the McLuhanian structural argument exist elsewhere. The operation of the British public-service broadcaster is laid down by a royal charter and a separate framework agreement, which describes the independence of editorial decisions together with the order of contact with the government — that is, contact is not prohibited but regulated and documented. In the German public-law model the supervision of broadcasters is carried out by bodies made up of delegates of social groups, whose composition the federal constitutional court has in several decisions narrowed in the direction of limiting political influence — here the guarantee lies in the structure of appointment. At EU level the European Media Freedom Act (EMFA) prescribes a transparent, non-discriminatory procedure and a public justification for the appointment and removal of the heads of public-service media, and it also extends to the publication of state advertising expenditure. The common feature of the three models is that none of them builds on individuals but on procedure and documentation — this is the international experience behind proposals 3.2 and 3.3.

Culture

  • KU2 — Media-pluralism monitoring
  • KU7 — Disinformation resilience and media-literacy programme

Transparency and anti-corruption policy

  • A5 — Whistleblowing system
  • A7 — Media pluralism as an institutional guarantee
  • A9 — Spin-dictatorship prevention index
  • A13 — Disinformation resilience

Digitalisation and AI regulation

  • D2 — Open data programme
  • D8 — Disinformation-detecting AI system

Proposed new programme point: An editorial statute and a complaints route outside the newsroom in public-service news provision — for the Culture area.

6.7 List of sources

Press sources (MIAK press monitor, 19 August 2026 — topic 1):

Knowledge base references (specialist books):

  • 📖 Sergei Guriev – Daniel Treisman: Spin Dictators
  • 📖 Marshall McLuhan: Understanding Media
  • 📖 Shoshana Zuboff: The Age of Surveillance Capitalism

MIAK internal materials:

  • MIAK policy area: Culture (background material and programme points; programme point ID: KU2, KU7)
  • MIAK policy area: Transparency and anti-corruption policy (programme points; programme point ID: A5, A7, A9, A13)
  • MIAK policy area: Digitalisation and AI regulation (programme points; programme point ID: D2, D8)
  • MIAK press monitor, 19 August 2026 — topic 1, score: 92/100

Supplementary public data sources:

  • European University Institute, Centre for Media Pluralism and Media Freedom — Media Pluralism Monitor indicator set
  • World Bank — Worldwide Governance Indicators, 2024
  • Documents on the application of the European Media Freedom Act (EMFA)
  • Reporters Without Borders — press freedom index

Generation metadata