← J·Index — The Journalism AI Index
J·Index / Countries / Hungary

AI in Hungary's news media

12 verified, source-linked AI implementations across 9 news organizations, 2024–2025.
Europe and Central Asia2024–2025
Cases
12
Organizations
9
Active since
2024
Controversies
2
Compared with the global pattern, Content Production & Automation accounts for 42% of cases here against 14% worldwide; and Research & Investigation accounts for 25% of cases here against 10% worldwide.

What AI is used for — and how that differs

Hungaryworldwide sharepp = percentage-point difference
Content Production & Automation5+27pp
Research & Investigation3+15pp
Editorial Workflow & Writing Assistance2+0pp
Verification & Fact-Checking1+1pp
AI Ethics & Editorial Policy1+4pp

How far adoption has gone

1 – Access0
2 – Adoption9
3 – Proficiency3
4 – Ways of Working0
5 – Reorganizing0

Who is adopting

Funding: Own budget (7), Public broadcaster / state media (4), Media dev org (1)

Notable cases

2025ATV

Journalist used AI to generate article about an Orbán parliamentary speech before it was delivered; article published at 9:09am describing a speech in past tense four hours before delivery (1pm); corrected at 11:30am; both the journalist and duty editor were dismissed; ATV declared ""texts written with artificial intelligence are not accepted

Content Production & Automation2 – Adoption⚑ controversy
AI text generation (unspecified)Source →

2025Direkt36

Used OpenAI and Google Gemini to extract structured data from 1,400+ scanned official government documents (hospital department closure decisions, 2020 onwards); OCR converted PDFs; AI extracted closure reasons, hospitals, durations; 5 editors verified and corrected AI-generated ~1,250-row database; investigation ""Kivérző kórházak"" revealed systemic collapse of Hungarian hospitals; published jointly with Telex

Research & Investigation3 – Proficiency
OpenAI / Google Gemini / OCR / PythonSource →

2025MUOSZ

Published joint declaration on AI in journalism (Nov. 10, 2025) signed by 14+ Hungarian independent outlets (444.hu, Magyar Hang, Népszava, Klubrádió, Magyar Narancs, Média1, and others); warns against AI manipulation in journalism and political campaigns; calls on newsrooms to refuse distributing AI-generated disinformation regardless of political leaning

AI Ethics & Editorial Policy2 – Adoption
Policy framework / joint declarationSource →

2024MTVA

Procured central AI services framework (Sept. 2024) covering image and audio processing, natural language text comprehension and translation, and partial or full workflow automation; framework agreement: 6.051 billion HUF over 60 months; first contract (459.9M HUF) awarded to Fornax ICT Kft.

Editorial Workflow & Writing Assistance2 – Adoption
AI services framework (contractors specify technology)Source →

2025M1 (MTVA)

Aired AI-generated footage in news broadcast (Déli Híradó) sourced from political organisation Nemzeti Ellenállás Mozgalom (NEM); included AI-generated person depicted as member of public criticising opposition tax plans; not labelled as AI-generated; widely criticised by independent media

Content Production & Automation2 – Adoption⚑ controversy
AI-generated video (tool unspecified;" produced by NEM)Source →

2024Lakmusz

Partner in EMIF-funded "The Check" multilingual AI fact-checking chatbot — RAG-based generative AI system ingesting verified fact-check database; Lakmusz provides Hungarian-language fact-checks; Newtral (Spain) led technical development; chatbot deployed across partner sites in Hungary

Verification & Fact-Checking2 – Adoption
Albania, LithuaniaSource →

Elsewhere in Europe and Central Asia

Albania 4Armenia 7Austria 22Azerbaijan 1Belarus 5Belgium 25Bosnia and Herzegovina 3Bulgaria 6Croatia 5Cyprus 1Czech Republic 24Denmark 27
Data from J·Index, the Journalism AI Index — every case verified against a published source. See the methodological note for how cases are collected and coded. Updated 2026-07-23.