Independent
by Exception

The 2026 edition of the State Media Monitor maps 588 active state and public media entries in 168 jurisdictions. Of these, 493, or 83.8 per cent, are controlled or captured by governments, ruling parties or interests tied to the state. The remaining 95, or 16.2 per cent, keep their editorial independence in one of the four independent forms defined by the State Media Matrix.

At first sight, 2026 brought the first improvement in the current five-cycle series. The non-independent share fell from 84.6 per cent in 2025 to 83.8 per cent, and the number of independent entries rose from 93 to 95. Both changes reflect the composition of the mapping. Reclassifications produced no net gain: six European entries gained independent status, while six entries elsewhere lost it. The net increase of two independent entries came from the addition of PBS and two Bolivian university broadcasters, offset by the removal of RNW Media after a scope review. Among the 574 entries mapped in both years, the independent count stayed at 92.

The fall in the non-independent share came mainly from Europe. There, six upgrades and the removal of nine Hungarian entries after the change of government (eight Captured Private outlets and the State-Controlled About Hungary) lowered the regional share from 53.4 to 43.8 per cent. Elsewhere, the trend ran the other way. Outside Europe, the non-independent share rose for the fourth consecutive cycle, to 92.5 per cent, its highest level since 2022. All six entries reclassified out of the independent family in 2026 are outside Europe: two in Indonesia and one each in Jamaica, Mexico, the Solomon Islands and the United Arab Emirates.

Independent state and public media are thus concentrated in a shrinking number of places. Europe accounts for 59 of the 95 independent entries and for 12 of the 18 Independent Public media, the category with the strongest protection from government interference. Only 47 of the 168 jurisdictions mapped have at least one independent state or public media entry, down from 51 in each of the 2022, 2023 and 2024 cycles. Across Sub-Saharan Africa, the Middle East and North Africa and Eurasia, which together account for 290 entries, six are independent.

The 2026 cycle also recorded more movement than any previous year. Forty-five entries changed category, against 16 to 21 in earlier cycles; 31 left the active mapping and 14 were added. Part of this movement comes from better evidence about who funds, owns and runs state media, which in Asia produced a large technical shift from State-Controlled to Captured Public classifications. Part comes from politics. In Hungary, the change of government altered the state links of previously captured media, prompting their removal from the mapping. In Indonesia, two public broadcasters upgraded in 2025 returned to the State-Controlled category a year later.

• Global

  493 of the 588 active entries (83.8 per cent) are non-independent, against 512 of 605 (84.6 per cent) in 2025. Independent entries rose from 93 to 95, with the whole net gain coming from additions to the dataset; Independent Public media held at 18, or 3.1 per cent of the total.

• Europe

  For the first time since 2022, independent entries form a majority of the European mapping: 59 of 105, or 56.2 per cent. Six public media were upgraded, including Spain’s RTVE and Czech Radio, and none was downgraded.

• The rest of the world

  Outside Europe, the non-independent share climbed from 90.6 per cent in 2022 to 92.5 per cent in 2026, rising in every cycle.

• Downgrades

  Six entries were reclassified out of the independent family, all outside Europe. Indonesia’s TVRI and RRI returned to State-Controlled one year after their upgrade.

• State-Controlled media

  The State-Controlled count fell from 392 to 367, largely because funding evidence moved 15 Asian outlets, among them Xinhua, Vietnam Television and the Vietnam News Agency, into the Captured Public category. Political control over them remains direct. Captured Public media reached a record 79.

• Sub-Saharan Africa

  Sub-Saharan Africa is the most tightly controlled region: 143 of 145 entries (98.6 per cent) are non-independent. All eight entries added in 2026 are State-Controlled, and the region now holds a third of the world’s State-Controlled media.

• Middle East and North Africa

  The independent share halved from 4.8 per cent in 2024 to 2.4 per cent, after Sky News Arabia was reclassified Captured Public. Only Israel’s Kan and Galatz remain independent.

• Eurasia

  58 of 60 entries (96.7 per cent) are non-independent. Five Azerbaijani outlets were removed because their ownership or control could not be verified; their removal does not demonstrate a reduction in state influence.

• Latin America and the Caribbean

  The non-independent share rose in every cycle, from 76.9 per cent in 2022 to 83.8 per cent. Autonomous universities host six of the region’s 11 independent outlets.

• United States

  Under the Trump administration, the U.S. Agency for Global Media remains State-Controlled. PBS returned to the dataset as Independent Public after a reassessment, despite the loss of its federal funding.

• Money as a lever

  Funding freezes, cuts and conditional budgets were a major instrument of pressure on independent public media, from Lithuania and Switzerland to Taiwan.

• Artificial intelligence

  In the regions reviewed for AI governance, published AI policies were most widespread among public media in Western and Northern Europe, with substantial gaps elsewhere. At most State-Controlled outlets, no published policy could be identified.

Table 1 sets the 2026 results against those of 2025. Both years are calculated from the State Media Monitor consolidated database 2022–2026, which records each entry’s classification in every cycle and retains removed entries in the historical record.

State-Controlled media remain the most widespread model of state media in the world. In 2026, 367 entries, or 62.4 per cent of the total, meet all three conditions of state control assessed by the State Media Matrix: the state funds them, controls their governance and steers their editorial line. Their geography is lopsided. Sub-Saharan Africa alone accounts for 122 of them, a third of the global total, and five regions, Sub-Saharan Africa, Asia, the Middle East and North Africa, Latin America and the Caribbean and Eurasia, together account for 338, or 92.1 per cent.

Captured media form the second tier of government influence. The 79 Captured Public or State Managed/Owned entries, 13.4 per cent of the total, are publicly owned or state-managed outlets whose editorial independence is compromised, while the full State-Controlled combination cannot be established, most often because predominant state funding cannot be demonstrated. Asia accounts for 28 of them, the Middle East and North Africa for 18 and Sub-Saharan Africa for 15. The 47 Captured Private entries, 8.0 per cent, are privately owned media whose editorial agendas are bent through sustained ties to governments, ruling parties or security establishments. They cluster in a handful of countries: Serbia and Turkey, which hold all 13 of Europe’s Captured Private media; Cambodia, with six; Iran and the Gulf states; Tanzania; and Russia.

The independent family comprises 95 entries. The most common independent form, with 50 entries, is Independent State-Funded and State-Managed media, which depend on the state for both money and management yet retain editorial control; 33 of them are in Europe. Independent Public media, insulated from government interference through both their funding and their governance, number 18. Fifteen entries are Independent State-Funded and 12 Independent State-Managed.

The regional contrasts are wide. In Sub-Saharan Africa, 98.6 per cent of entries are non-independent, in the Middle East and North Africa 97.6 per cent and in Eurasia 96.7 per cent. Asia follows with 91.9 per cent and Latin America and the Caribbean with 83.8 per cent. Europe, at 43.8 per cent, is the only large region where independent entries form a majority. North America and Oceania, with four and ten entries respectively, are the regions with the highest independent shares, though each has had an entry reclassified as State-Controlled in the past two cycles.

Five consecutive annual cycles now allow the State Media Monitor to separate durable trends from yearly noise. In the consolidated database, the number of active entries stood at 595 in 2022 and 2023, rose to 605 by 2025 as coverage expanded in Sub-Saharan Africa, the Caribbean and Oceania, and fell to 588 in 2026, when exits outnumbered additions for the first time. The non-independent share climbed from 82.5 per cent in 2022 to 84.6 per cent in 2025 before easing to 83.8 per cent. The independent family shrank from 104 entries to 93 over the first four cycles and recovered slightly to 95.

The composition of non-independent media changed more than its size. State-Controlled media peaked at 393 in 2024, stood at 392 in 2025 and fell to 367 in 2026, their lowest level on the series. Captured Public media rose from 67 in 2025 to 79 in 2026, compared with 59 in 2022. Part of that shift is technical: in Vietnam and China, departmental accounts and municipal financial records allowed the funding condition to be tested more precisely, and outlets for which predominant state funding could not be demonstrated moved from State-Controlled to Captured Public. Combined, captured public and captured private media rose from 108 entries in 2022 to 126 in 2026, or from 18.2 to 21.4 per cent of the total. Some of this growth reflects better evidence about the ownership and funding of outlets already mapped.

At the top of the scale, the picture barely moved. Independent Public media numbered 20 in 2022 and 2023 and have stood at 18 since 2024, roughly 3 per cent of all entries. Behind that stability lies constant movement. Czech Radio left the category in 2023 and returned in 2026; Slovenia’s RTV Slovenija joined it in 2023; South Korea’s KBS dropped out in 2024 and returned in 2025; Thailand’s Thai PBS left in 2024 and Lithuania’s LRT in 2026.

The regional series diverge sharply. Europe is the only region where the independent share has grown markedly over the period, with the entire gain occurring in 2026. It fell from 51.3 per cent in 2022 to 47.1 per cent in 2024 and 46.6 per cent in 2025, as Spanish regional broadcasters, the Greek news agency, Czech Radio and Albania’s RTSH lost independent status and captured Hungarian outlets were added, before jumping to 56.2 per cent. The non-independent share is higher in 2026 than in 2022 in every other large region except Eurasia, where it has hovered near 97 per cent throughout. In Latin America and the Caribbean it rose in every cycle, from 76.9 to 83.8 per cent. In the Middle East and North Africa it moved from 95.4 to 97.6 per cent, and in Sub-Saharan Africa from 97.6 to 98.6 per cent. Asia’s share swung between 90.8 and 94.4 per cent as outlets moved in and out of the independent family. Taken together, the non-independent share outside Europe rose from 90.6 per cent in 2022 to 92.5 per cent in 2026, without a single year of decline.

The map of independence has narrowed accordingly. In 2022, 51 jurisdictions had at least one independent state or public media entry; in 2026, 47 do. Over five cycles, 11 jurisdictions lost their last independent entry, among them Greece, Indonesia, Iran, Thailand and Vietnam, while seven gained one, including Croatia, Japan, North Macedonia and Slovenia.

Europe produced the year’s most visible improvement. Of its 105 active entries in 38 jurisdictions, 59, or 56.2 per cent, belong to the independent family, the first independent majority since 2022. The roster shrank by 11 entries, nine of them Hungarian outlets removed after the parliamentary election of 12 April 2026, in which the Tisza Party won 141 of the 199 seats and ended 16 years of Fidesz-KDNP rule. The change of government altered the state links of these previously captured outlets, prompting their removal; none of them became an independent entry. Hungary’s public broadcaster remains State-Controlled on a provisional basis, and whether it will produce independent output under the framework adopted by Parliament in June has yet to be assessed.

Six public media moved into the independent family. Spain’s RTVE, Aragón’s CARTV, Bulgarian National Radio, Croatia’s news agency HINA and North Macedonia’s MRT moved from State-Controlled to Independent State-Funded and State-Managed, and Czech Radio returned to the Independent Public category. Several of these upgrades are provisional because appointment by parliamentary majorities and dependence on annual budgets leave them exposed.

Funding was the main source of pressure on Europe’s strongest public broadcasters. Lithuania’s LRT dropped from Independent Public to Independent State-Funded after a law froze its budget at EUR 79.6 million a year for 2026 to 2028. Swiss voters rejected a cut of the broadcasting levy in March, yet SRG SSR still plans to cut up to 900 full-time-equivalent positions and save CHF 270 million by 2029, while Finland’s Yle and Ireland’s RTÉ face their own shortfalls. Capture persists where governments control appointments: in Slovakia, where the government and governing parties received 71.3 per cent of political coverage in STVR’s main news in a February monitoring period; in Romania, where interim heads proposed by AUR and PSD took over TVR and SRR; and in Serbia and Turkey, home to all of Europe’s Captured Private media. In Serbia, the Student List won the early parliamentary election of 25 October 2026, held after the review period, and Aleksandar Vučić conceded, opening the next test for the region.

Eurasia remains the most uniformly controlled region outside Africa and the Middle East. Of its 60 entries in 12 jurisdictions, 58, or 96.7 per cent, are non-independent: 48 State-Controlled, six Captured Public and four Captured Private. Only Moldova’s Teleradio-Moldova and Ukraine’s Suspilne sit in the independent family, two public broadcasters at the region’s western margin whose governance remains exposed to political and security pressure. Russia, Central Asia and the South Caucasus contain no independent entry at all.

The roster fell from 69 entries to 60 through nine exits and no additions. None of these removals established editorial independence. Five Azerbaijani outlets, among them Space TV, ATV and Real TV, were removed because no authoritative, current record of their owners or of continuing state control could be obtained. Evidence of their alignment with the government remains, and their removal does not demonstrate a reduction in state influence. The removals coincided with an intensifying campaign against the remaining independent journalism in the country: in the Toplum TV case, nine defendants received prison terms totalling about 121 years, and in April 2026 the Supreme Court upheld sentences totalling 59.5 years against six Abzas Media journalists and RFE/RL journalist Farid Mehralizada. Azerbaijan’s mapped set now represents the verifiable minimum of state-controlled media in the country.

In Russia, four entries, National Media Group, Tsargrad TV, Oriental Review and African Initiative, moved into the Captured Private category, where private ownership structures coexist with documented state-linked editorial control. Moldova’s news agency Moldpres moved from State-Controlled to Captured Public. Uzbekistan’s reform rhetoric has produced new outlets and modernised branding, while all five of its mapped entries remain non-independent.

In the Middle East and North Africa, 83 of 85 entries in 19 countries, or 97.6 per cent, are non-independent. The region counts 52 State-Controlled entries, unchanged from 2025, together with 18 Captured Public and 13 Captured Private entries. At 36.5 per cent, it has the highest regional share of captured media. Its two independent entries are both Israeli: the public broadcaster Kan and the military radio station Galatz, which have so far withstood sustained government pressure.

The direction of travel in 2026 was uniform. Two Iranian outlets linked to the Revolutionary Guards, the Tasnim News Agency and Sobh-e Sadegh, entered the mapping as Captured Private, and the Syrian Arab Publishing and Distributing Company, Captured Private since 2022, left it. Sky News Arabia, previously Independent State-Managed, was reclassified Captured Public, its ownership and editorial alignment placing it within the Abu Dhabi state-media orbit. With Iran’s ISNA reclassified in 2025, the region’s independent share has halved in two years, to 2.4 per cent.

The region’s signature is the Gulf media conglomerate: commercially sophisticated, internationally distributed and often popular groups such as Saudi Arabia’s MBC and SRMG, the United Arab Emirates’ Abu Dhabi and Dubai networks and Qatar’s Al Jazeera, whose ownership, financing and red lines keep them aligned with their governments. This explains why the State-Controlled share in the region, 61 per cent, is far lower than in Sub-Saharan Africa while overall non-independence remains similarly high. Commercial financing and corporate structures coexist with continued political control. Syria’s tentative opening after the fall of Bashar al-Assad remains the only development pointing the other way.

Sub-Saharan Africa is the most tightly controlled region in the dataset. Of 145 entries in 41 countries, 143, or 98.6 per cent, are non-independent, and 122 are State-Controlled, up from 116 in 2025. The only independent entries are two Ivorian outlets classified Independent State-Funded and State-Managed, the publishing house SNPECI and the news agency AIP, exactly as in 2025.

The dataset grew by eight entries, every one of them State-Controlled: news agencies, government publishers and ministry-linked broadcasters in Ethiopia, Malawi, Mozambique, the Republic of the Congo, Sierra Leone, The Gambia and Burundi. Two outlets, Eritrea’s Tesfa News and Ethiopia’s Fana Media Corporation, were reclassified from State-Controlled to Captured Private, joining four ruling-party-linked or state-dependent conglomerates in Tanzania. Their formal ownership sits outside the state, while their editorial lines continue to serve those in power.

Apparent gains in public broadcasting rest on fragile foundations. South Africa’s SABC has reduced its debt and earned praise for more balanced coverage, but its licence-fee base has collapsed, and funding reforms have repeatedly stalled. In Nigeria, the leadership of the Nigerian Television Authority remains a presidential appointment. Chinese state media continue to deepen their presence across the continent through content partnerships, satellite television and training programmes for African journalists, reinforcing the dependence of already state-aligned outlets.

Asia recorded the largest movement between categories in any region in 2026, and little change in substance. Of 111 entries in 26 jurisdictions, 102, or 91.9 per cent, are non-independent, against 91.2 per cent in 2022. State-Controlled entries fell from 86 to 64 and Captured Public entries rose from 15 to 28, mainly because 15 outlets, eight in Vietnam, four in China and one each in Nepal, Malaysia and Thailand, moved from State-Controlled to Captured Public. The move reflects new financial evidence: departmental accounts and restructuring records showed that predominant state funding could not be demonstrated for these outlets. Their political control remains direct.

The changes that mattered most for independence ran in one direction. Indonesia’s TVRI and RRI, upgraded in 2025, returned to State-Controlled. In June 2026, RRI’s president director told parliament that the broadcaster is an instrument of the state whose programmes support the government’s strategic agendas; in August, a deputy communication minister directed TVRI to translate state policy into its editorial policy, programmes and content. No Asian outlet entered the independent family, and the independent count fell from 11 to nine. Independence is now confined to four of the 26 jurisdictions: South Korea, whose KBS, MBC, EBS and Yonhap are Independent Public; Taiwan, with three Independent State-Funded and State-Managed entries; Japan, where NHK is Independent State-Managed; and Bhutan, whose Kuensel is the only independent entry in South Asia. All 50 entries in Southeast Asia are non-independent.

Even this small independent group is under strain. South Korea is implementing a 2025 governance reform for KBS, MBC and EBS that broadens board nominations, though implementation ran well beyond the statutory timetable. In Taiwan, the Legislative Yuan made TWD 200 million of public television funding conditional on the appointment of a chairman for the new PTS board, which has not been constituted since May 2025. In Bangladesh, the reform opening identified in 2025 did not materialise, and the national broadcasters and news agency remain State-Controlled.

Latin America and the Caribbean combine competitive media markets in many countries with weak safeguards for state-funded broadcasting. Of 68 entries in 25 countries, 57, or 83.8 per cent, are non-independent, 52 of them State-Controlled. The non-independent share has risen in every cycle since 2022, when it stood at 76.9 per cent.

Two established national services were reclassified out of the independent family in 2026. The review assessed the Public Broadcasting Corporation of Jamaica against the 2010 amendment regulations that govern it and moved it from Independent State-Funded to State-Controlled. Most of its board members are appointed by the Governor-General after consultation with the Prime Minister and the Leader of the Opposition, with no independent nomination power, protected terms or statutory editorial firewall. Its current chair also serves as a government senator. Mexico’s indigenous-language radio network SRCI also became State-Controlled. The only additions to the independent family were two Bolivian university broadcasters.

Universities provide the region’s most reliable route to independence: six of its 11 independent outlets are housed in autonomous public universities, from Mexico’s TV Radio UNAM to Costa Rica’s university radio and television system. The principal national broadcasters tell a different story. With the exception of Chile’s Captured Public TVN, every main national state broadcaster in South America is State-Controlled. Cuba, Nicaragua and Venezuela run comprehensively controlled state media systems, yet State-Controlled national broadcasters also operate in Uruguay, Brazil, Peru, Panama, Paraguay, Barbados, the Bahamas and Jamaica, several of which rank among the freest media environments in the Americas. In the region, competitive elections and pluralistic private media have rarely translated into arm’s-length governance of the national broadcaster.

In the United States, the Trump administration is mounting an unprecedented attack on independent media, with public broadcasting among its principal targets. The U.S. Agency for Global Media, downgraded in 2025, remains State-Controlled. The classification rests on reported intervention in Voice of America’s editorial decisions under government-appointed management. A federal court ruled in March 2026 that Kari Lake had served unlawfully as the agency’s acting chief executive, and Congress appropriated USD 652.7 million for the 2026 fiscal year, yet by October VOA was broadcasting in only eight languages. PBS, by contrast, returned to the dataset as Independent Public. It had been left out in 2025 after Congress rescinded federal funding for public broadcasting, which led the Corporation for Public Broadcasting to dissolve itself in January 2026. The 2026 reassessment applied the same tests used for NPR: continuing state funding is not a condition of the Independent Public model, and PBS’s governance and editorial safeguards were unchanged. Canada’s CBC remains Independent State-Funded.

In Oceania, seven of the ten entries are independent, led by the public broadcasters of Australia and New Zealand. The region’s three State-Controlled entries are all in small Pacific states. The 2026 review reclassified the Solomon Islands Broadcasting Corporation from Independent State-Funded and State-Managed to State-Controlled, citing its placement directly under the Office of the Prime Minister and Cabinet since 2022, its reliance on an annual government subvention and censorship instructions documented in 2022 and 2026 under two administrations. A change of government in May 2026 offers an opening for reform. In Samoa, the government newspaper Savali was added as State-Controlled.

Several of the largest changes of 2026 followed shifts in political power. In Hungary, the change of government after the April election altered the state links of previously captured media, prompting the removal of nine entries from the mapping, while the public broadcaster remains provisionally State-Controlled. South Korea’s governance reform of KBS, MBC and EBS followed the 2025 change of government. In the Solomon Islands, a new government took office in May 2026, with an opportunity to revisit the arrangement that has placed the public broadcaster under the Prime Minister’s office since 2022. In Serbia, the election of 25 October may open a similar chapter.

Change that arrives with a new government can leave with the next one unless safeguards constrain those who hold power.

Political change offers no guarantee of reform. In Romania, the leadership of public radio and television passed to interim heads proposed by political parties. In Bangladesh, the reform commission set up after the fall of Sheikh Hasina proposed autonomous public media structures, but none of its principal recommendations had been enacted by the 2026 review. Change that arrives with a new government can leave with the next one unless safeguards constrain those who hold power.

The 2026 downgrades show the gap between formal safeguards and effective protection. Indonesia’s TVRI and RRI operate under regulations that define them as independent and prohibit outside interference in their operations, and RRI’s content is overseen by the independent Indonesian Broadcasting Commission. Those provisions did not prevent a deputy minister from directing TVRI to make state policy its editorial policy, or RRI’s leadership from describing its programmes as support for government agendas. In Jamaica and the Solomon Islands, the 2026 review found that governance arrangements in place since 2010 and 2022 respectively offered no effective protection against political influence over appointments or content.

Where independence has proved durable, formal guarantees are backed by institutional design. In Latin America, constitutional university autonomy shields six of the region’s 11 independent outlets. In Europe, Article 5 of the European Media Freedom Act has since August 2025 required transparent appointment procedures and adequate, predictable funding for public service media, and the European Commission sent Hungary a letter of formal notice covering public service media in December 2025.

Governments use budgets to exert pressure on independent public media. Lithuania has frozen LRT’s funding for 2026 to 2028, suspending the statutory formula that had protected it. Taiwan’s legislature tied part of public television’s budget to a board appointment controlled by political institutions. A Czech government bill proposes to replace the licence fee of Czech Television and Czech Radio with a lower allocation from the state budget. Pakistan abolished the licence fee of Pakistan Television and moved it onto federal grants, a change that took the broadcaster into the State-Controlled category.

In the United States, Congress rescinded federal funding for public broadcasting at the Trump administration’s request. This led to the dissolution of the Corporation for Public Broadcasting, forced PBS to cut its budget by 21 per cent and left NPR facing layoffs. Where funding pressure has not compromised editorial independence, as in Switzerland, Finland and Ireland, it is steadily narrowing what public broadcasters can do.

Captured media account for a growing share of the mapping. Captured public and captured private media together rose from 108 entries in 2022 to 126 in 2026, more than a fifth of the total, though part of the increase reflects better evidence about the ownership and funding of outlets already mapped. The structures vary: Revolutionary Guards media groups in Iran, ruling-party conglomerates in Tanzania, commercial groups in Serbia and Turkey, privately owned outlets aligned with the government in Cambodia, and the Gulf’s large commercial broadcasters. In 2026, reassessments of ownership and control moved four Russian entries and two in the Horn of Africa into the Captured Private category. In Turkey, the Can Medya group passed under the management of a state deposit insurance fund and became Captured Public.

China’s expanding media footprint extends state influence across borders. Content partnerships with Xinhua and China Media Group, funding for local outlets and training programmes for journalists from Africa and Southeast Asia can deepen the dependence of already state-aligned media on another state. In 2026, the Chinese New Zealand Herald was reclassified Captured Private after corporate research confirmed private ownership, while evidence of sustained Chinese state-media influence remained.

In such systems, the mapping captures only the state influence that can be verified.

In several closed systems, entries were removed because state control could not be documented to the Monitor’s evidentiary standard. Five Azerbaijani outlets left the mapping because their current owners or continuing state control could not be established, at a time when courts handed long prison sentences to independent journalists. In Russia, Rambler&Co was removed for the same reason. In Vietnam, the consolidation of municipal press organisations into integrated agencies reduced the count, while political and editorial control over the merged agencies remains direct. In such systems, the mapping captures only the state influence that can be verified.

Editorial governance of artificial intelligence advanced quickly in 2026. The 2026 reviews of Europe and Asia examined published AI policies outlet by outlet, and found them most widespread among public media in Western and Northern Europe, with substantial gaps elsewhere. Germany’s ARD, ZDF, Deutschlandradio and Deutsche Welle adopted a joint AI code in January 2026, SRG SSR’s AI regulations took effect in June, and Radio France banned synthetic voices that could replace the vocal identity of its journalists and presenters. Elsewhere in Europe, published rules are concentrated in Spain, Italy and Portugal, and remain rare in Central and Eastern Europe and Turkey. In Asia, publicly accessible rules exist at Japan’s NHK, South Korea’s KBS, MBC and Yonhap, Taiwan’s PTS and Central News Agency, Thai PBS and Singapore’s Mediacorp. In the United States, NPR and PBS have had guidance on generative AI since 2023.

In the systems with the tightest political control, the rules governing AI in newsrooms are written by the same state that directs those newsrooms.

The principles recur across these documents: human editorial responsibility, verification before publication, protection of confidential material and labelling where AI plays a material role in content. Luxembourg’s 100,7 explicitly prohibits the use of generative AI to establish or verify facts. At most State-Controlled outlets, by contrast, no published AI policy could be identified, and in the most closed information environments even the extent of newsroom use could not be established. China Media Group has adopted a formal usage standard, but its full rules are not public.

Regulation is moving in parallel. The transparency obligations in Article 50 of the EU AI Act have applied since 2 August 2026. Deployers must disclose deepfakes, and AI-generated or manipulated text published to inform the public on matters of public interest, unless the text has undergone human review or editorial control and a natural or legal person holds editorial responsibility for its publication. South Korea’s and Taiwan’s AI basic acts took effect in January 2026. In Vietnam, the AI law in force since 1 March 2026 requires machine-readable marking of AI-generated audiovisual material, and Decree 237/2026/ND-CP, implementing the Press Law, requires press organisations to label AI-generated content that could mislead audiences and to verify and take responsibility for it. China has applied national labelling rules for synthetic content since September 2025. In the systems with the tightest political control, the rules governing AI in newsrooms are written by the same state that directs those newsrooms.

The 2026 State Media Monitor records a modest global improvement that reflects changes in the composition of the mapping, concentrated in Europe. Outside Europe, the share of state and public media under government control or capture rose for the fourth consecutive cycle, to 92.5 per cent. Globally, 47 jurisdictions have at least one independent state or public media entry, down from 49 in 2025 and 51 in 2022. Across Sub-Saharan Africa, the Middle East and North Africa and Eurasia, independent state or public media exist in only four countries: Côte d’Ivoire, Israel, Moldova and Ukraine.

Three questions will shape the next cycle. The first is whether Europe’s gains hold. Hungary’s public broadcaster has yet to show that its new framework produces independent output, Serbia faces a transition after the October election, and several of the 2026 upgrades remain provisional. The European Media Freedom Act provides the benchmark; its enforcement will decide whether those requirements become durable national safeguards or commitments revisited after every election. The second is money. From Lithuania and Switzerland to Taiwan and the United States, independent public media face frozen, conditional or shrinking budgets, a form of pressure that narrows their capacity even where editorial independence holds. The third is opacity. Where ownership and control are hidden behind nominally private vehicles and commercial conglomerates, capture becomes harder to document, and the State Media Monitor’s counts in closed systems describe a verified minimum.

The conditions that sustain editorial independence are well known: insulated and predictable funding, appointments removed from political control, and editorial safeguards that are enforced in practice. Only 18 entries, 3.1 per cent of the total, meet the Independent Public standard of insulation through both funding and governance.

The lesson of 2026 is that editorial independence can change quickly, while durable protection requires safeguards that constrain political power in practice. Legal guarantees must be supported by independent appointments, predictable funding and effective enforcement.

The State Media Monitor is built on the State Media Database, first created by Marius Dragomir in 2004 and updated annually on a global basis since 2020. The 2026 cycle maps 588 active entries in 168 jurisdictions, based on documentary research and interviews with local media experts, journalists and civil society representatives.

Each entry is assessed against three criteria: funding, ownership and governance, and editorial control. On that basis, the State Media Matrix assigns it to one of seven categories. State-Controlled media meet all three conditions of state control. Captured Public or State Managed/Owned media are publicly owned or state-managed entries whose editorial independence is compromised, where the full State-Controlled combination is not established. Captured Private media are privately owned outlets whose editorial agendas are captured through sustained political, state or ruling-party relationships. The four independent categories, Independent Public, Independent State-Funded, Independent State-Funded and State-Managed, and Independent State-Managed, describe entries that maintain editorial autonomy under different forms of public funding, ownership or management. Since 2022, the study has treated each media organisation as a single entity across television, radio, print and online platforms, regardless of how many channels or titles it operates.

Evidence is drawn from three tiers of sources: annual reports, financial statements and legal acts; media coverage, NGO reports and academic studies; and interviews with experts and insiders, investigative reporting and, in selected countries, content analysis during politically sensitive periods. Reclassification and removal follow the evidence available in each cycle. New accounts, appointment records, corporate filings or a change of government can change an entry’s category, or lead to its removal, without any corresponding change in editorial behaviour. Removal for lack of verifiable ownership or control evidence does not imply independence.

The 2026 classifications reflect the evidence available at the close of each regional review in 2026. Later developments, such as Serbia’s election of 25 October 2026, are mentioned in the text and will be assessed in the next cycle.

Historical figures in this study are calculated from the State Media Monitor consolidated database 2022–2026 (version of 1 November 2026). The database records each entry’s classification in every cycle, including retrospective corrections, and retains removed entries in the historical record, recording when and why they left the mapping. All tables and charts are calculated from it. The figures may differ from those published in earlier editions and in some regional overviews, which present totals as recorded at the time. The 2025 edition, for example, reported 606 entries, of which 512, or 85 per cent, were non-independent; the consolidated database records 605 active entries for the 2025 cycle, of which 512, or 84.6 per cent, were non-independent.