When Propaganda Backfires: Lessons from Orban’s Social Media Defeat in Hungary Election

It was clear from the heavy defeat of Viktor Orban’s government in April's election that its overwhelming dominance of Hungary’s media landscape wasn’t enough to get it across the line this time. Since then, emerging data shows just how ineffective the election campaign of the former Hungarian prime minister’s Fidesz party, its Christian-nationalist KDNP ally and its proxies really was. According to the Vilnius-based narrative intelligence company Repsense, which combined three complementary analyses in the four months leading up to Hungary’s general election, several of the government’s efforts, including a false-flag operation involving an actual flag to discredit the opposition Tisza party, backfired badly. As detailed in its non-public consolidated executive report seen by BIRN, Repsense tracked a total of 816,863 articles and posts across Hungarian-language news sites and Facebook, X, YouTube, TikTok, Instagram and Telegram between December 2025 and April 2026, containing 12.25 million unique sentences. (The company does not attempt to capture every account, but focuses its data collection on influential accounts and widely seen content across the political spectrum.) The report shows that while content matching Fidesz campaign themes outweighed that matching opposition campaign themes by roughly four to one in volume, in several cases the counter-response travelled further and faster than the initial operation itself. That doesn’t mean the propaganda of the former government didn't reach large numbers of Hungarians. After 16 years in power, Fidesz directly or indirectly controlled an estimated 80 per cent of Hungary's media landscape, spanning public and commercial broadcasters, national dailies, regional press and online outlets. The government's messaging was further reinforced by ubiquitous nationwide billboard campaigns. However, convincing their citizens proved a different story this time around. Two young men unfurl and hold up the Ukrainian flag at the March 15 national holiday demo of the Tisza party in Budapest, Hungary. Screenshot: YouTube / Telex.hu Debunkings went more viral A notable case was when several young people unfurled a huge Ukrainian flag at the Tisza party’s March 15 national holiday rally. Photos of the protest action were taken and rapidly spread by pro-Fidesz outlets and accounts, portraying the incident as evidence of the opposition party’s pro-Ukrainian alignment. However, the independent outlet 444.hu quickly reported that the protest action had been coordinated by key staff members of a Fidesz-linked digital campaign agency. It also identified the two minors who had unfurled the flag, one of whom later that day posted an Instagram story showing him with the chairman of the Budapest organisation of Fidelitas, Fidesz’s youth wing. Repsense examined the content database’s dimensions like viral intensity, which means how fast and how far a narrative spreads – a measure “driven mainly by posting speed (sentences per hour), acceleration, and how big the narrative got in its first six hours relative to its final size”. Regarding the flag issue, the original posts reached 28 points on their “Facebook deep-dive” analysis’ viral intensity scoring, while the exposure scored 44, achieving half of its total reach in under an hour – the fastest spread in their Facebook dataset. “What the data shows is that Hungarian citizens used the same platforms not as passive consumers of propaganda, but as an active counter-disinformation force. This is a dynamic that echoes the earliest promise of social media as a tool for civic resistance, which brings back memories of the Arab Spring,” says Mykolas Katkus, co-founder and CEO of Repsense. “The information environment had developed an immune response – each new provocation only accelerated its own exposure.” That the truth behind the flag-unfurling was exposed by an independent media outlet highlights the importance of the free press. “I believe that the report is not fully holistic in the sense that we couldn’t make this government change without independent journalists,” emphasises Kata Horvath, a media researcher at the Department of Media and Communication of Eotvos Lorand University in Budapest. Horvath believes Hungary’s remaining free press did a tremendous job under very trying conditions, uncovering many corruption cases, the truth behind the flag incident and the paedophile scandal. The latter refers to the 2024 resignation of president Katalin Novak after it emerged that she issued a pardon to the former deputy director of a children’s home who had helped cover up child sexual abuse by his superior. Bulcsu Hunyadi, head of programs at the Budapest-based Political Capital think-tank, says the active communication by trusted and popular experts also proved important in explaining to the public what was really going on. Repsense’s data shows similar patterns regarding the divergence between independent pollsters showing the opposition’s advantage and Fidesz-aligned institutes that predicted the governing party would win again. This issue became a narrative thread itself; in the report’s comparison, a poll by the Fidesz-aligned Nezopont Institute scored 15 on the viral intensity index, while the counter-response scored 35. An AI-generated video featuring a Hungarian father being executed, part of Fidesz’s efforts to frame the vote as a choice between staying out of the war in Ukraine or joining it, scored 28 points, while its counter-narratives got 35. Repsense found that this was one of the most negatively received pieces of campaign-produced content. While these negative reactions could have been towards Ukraine or simply a human reaction, the surrounding discourse, rather than the sentiment score itself, suggests a backlash against the video itself. The most negatively received narrative threads, however, concerned physical attacks on activists. The pattern was not universal, however: in the widely discussed case of a Ukrainian cash-in-transit seizure that was heavily used for propaganda purposes by the pro-government media, Repsense measured the original narrative and the counter-narrative at the same viral intensity. A Fidesz supporter holds a placard in Kossuth Square in Budapest, on 15 March 2026, with the inscription: “Say No to the Ukrainian puppet”. Photo: Alexander Faludy Familiar narratives Analysing 12.25 million sentences collected from news sites and social media platforms, Repsense found that 2.81 million sentences matched Fidesz campaign narratives, almost 23 per cent of the examined content. Three framings – the accusation that the opposition Tisza party wanted to drag Hungary into war, that its would wreck the economy, and that it is funded by Ukraine and other foreign powers – accounted for nearly 20 per cent of the tracked discourse. According to the investigative outlet Vsquare.org, Moscow dispatched social media manipulation specialists to Budapest to interfere in the election in favour of Orban’s Fidesz. As Repsense’s data highlights, the distribution of this investigative report became a major narrative in itself, spread rapidly and organically by authentic opposition Facebook groups rather than bots. Repsense’s analysis confirms that Kremlin narratives – Orban as a guarantor of peace, Brussels/Ukraine as a war threat, Magyar as Brussels’ puppet, Ukraine as a meddler in Hungary’s election – “were clearly present”, but noted they tended to be amplified mainly through official Fidesz party accounts and pro-government media outlets. The Kremlin narrative playbook was clearly measurable on TikTok, accounting for around 17 per cent of clustered posts, with Fidesz-aligned accounts among its leading carriers. However, the Repsense report stresses that its “attribution to Russian coordination cannot be established from content data alone”. In any case, it’s hard to tell to what extent Fidesz invented these talking points on its own. Repsense did note that a Russian-linked bot network spread fabricated assassination-attempt videos about Orban from February onwards. Katkus says the situation on TikTok is different to that on Facebook. “An average person who just registered on TikTok in Budapest will see something that is strongly biased,” he emphasises. In a separate TikTok analysis, Repsense examined 15,791 unique posts from a dataset of 183 accounts. Monthly posting volume among these accounts rose from 488 posts in December 2025 to 7,843 in March 2026, a roughly 16-fold increase. Daily posting activity among the monitored accounts then peaked at 360 posts on April 11, the day before the election. The researchers also recorded spikes in posting activity among these accounts on March 3-4, coinciding with an Orban-Putin phone call and a Putin-Szijjarto meeting in the Kremlin, and again on April 8, following a leaked-audio story involving Szijjarto. The report stresses that the timing shows correlation rather than proof of coordination, but says it is “consistent with the documented campaign timeline”. Katkus points out that “Russians are exporting their own mindset”, so we can see similar tactics employed in different countries; however, it is “frequently not effective”. One of these is to frame the vote geopolitically instead of around bread-and-butter issues, which was originally a Russian internal tactic. “They are trying to export some sort of black-and-white worldview,” he says, adding that the villain doesn’t necessarily have to be Ukraine, it can also be NATO or another neighbouring country, sometimes using almost identical phrasings. Other tactics, he says, involve discrediting the entire voting process and attacking anti-Russian candidates. Leader of the Tisza Party, Peter Magyar, addresses his supporters during the National March in Budapest, Hungary, 15 March 2026. EPA/Tibor Illyes What changed in Hungary? After returning to power in 2010, Viktor Orban claimed that Fidesz had won the election online and on Facebook. Since then, however, Fidesz has frequently referred to social media as a new and hostile terrain where leftist and liberal voices predominate. One of the pro-Fidesz camp’s major social media projects, launched in 2020 ahead of the 2022 general election, was the “Megafon Centre”, which trained pro-Fidesz content creators to produce short-form videos and heavily promoted their content through paid advertising. By late 2025, Megafon had spent more than 2.7 billion forints (about 7.5 million euros) on Facebook advertising alone. While a ban on political ads by Meta and Google in the autumn of 2025 made it harder to flood social media with propaganda, Fidesz and their proxies didn’t stop trying. Recently, the former government’s campaign founded groups called “Fight Club” (Harcosok klubja) which focused on winning online battles, and “Digital Civic Circles” (Digitalis Polgari Korok) to build communities. “I think it’s safe to say that this was a huge failure,” says Horvath of Eotvos Lorand University. She explains this by highlighting the grassroots nature of social media, arguing that ideas cannot simply be pushed from the top down. Meanwhile, Hungarian fact-checking outlet Lakmusz and Political Capital have repeatedly documented networks of fake Facebook profiles and other forms of coordinated social media manipulation by Fidesz. Horvath emphasises that while these networks clearly put considerable effort into these methods, they didn’t work well, partly “because it wasn’t authentic”. Not only did these social media campaigns fail, but so too did Fidesz’s broader propaganda apparatus, which deployed the same smear tactics against Magyar that it had used against previous challengers. This time, however, the accusations failed to stick because, Horvath argues, Hungary now had a viable opposition party. She adds that Fidesz’s campaign started looking really tired, not least since its “pro-family” stance was fatally undermined by the paedophile scandal. Horvath says this debacle and the rise of Magyar meant Fidesz wasn’t able to silence people anymore. Bulcsu Hunyadi, head of programs at Political Capital, also dates the erosion of trust back to then, but cites other reasons too. “The economic situation and deteriorating public services also paved the way. The underlying impression of many people was that their economic and social situation, as well as that of the country, was deteriorating,” he says. Hunyadi points out Fidesz’s popular support was based on the promise that if you overlook the party’s ideology, anti-democratic measures and corruption, you could benefit from it. Other scandals also challenged Fidesz’s credibility in key areas, like the health and environmental effects of the battery plant near God; Orban’s endorsement of George Simion, a Romanian far-right candidate with a history of provocations against ethnic Hungarians; and the misuse of money channelled into the foundations of the Hungarian central bank. He concludes that Repsense’s study proves Fidesz’s propaganda and disinformation campaigns in this election campaign failed because of the party’s falling popularity and the emergence of a challenger who offered alternative narratives about the country and who increasingly looked as though he would win. “It was not primarily the failure of disinformation that caused the regime’s downfall, and disinformation wasn’t the only reason why it worked for 16 years,” Hunyadi tells BIRN. Then-PM Viktor Orban posing for a photo with supporters during a rally at a campaign stop in Salgotarjan, Hungary, 11 April 2026. EPA/HUNGARIAN PM GENERAL DEPARTMENT OF COMMUNICATION / AKOS KAISER Laughter is dangerous After Fidesz and its KDNP ally secured successive election wins with a constitutional majority, it seemed as though they would never be voted out of office. Yet with Magyar’s rapid rise, this perception changed, causing what Hunyadi calls a “social awakening”. Hungarians began talking about politics and participating in demonstrations more frequently. This vibe shift was even clearer at the end of the campaign, when several serving and former public officials and public servants spoke out about the government’s abuses of power and ineffective management of public services. According to Repsense, the March 15 national holiday, with 19,728 Facebook posts, generated more public debate than the preceding three weeks combined, including descriptions of the Tisza party’s “historic” mobilisation for its demo in Budapest. “From this point on, the discourse treated a Tisza victory as a working assumption,” the report says. Orban’s main challenge was that by the end of the campaign, Hunyadi says, his messages didn’t spark anger so much as laughter. Being ridiculed was a threat that Orban’s campaign team took seriously. In a closed group, governing politicians reportedly asked party members to like their public posts to combat oppositional “attacks”, meaning the high numbers of laughing emojis attached to them. On Facebook, there were a huge number of parody pages and groups sprang up around Fidesz’s Digital Civic Circles initiative. Repsense’s Katkus says that while debunking disinformation works in some places like Moldova, on social media it can mean bringing “more attention” to the topic, so sometimes laughing at propaganda can be more effective. Hunyadi stresses that the gap between the government propaganda’s parallel reality and the actual situation Hungarians experience on a day-to-day basis “significantly increased”. According to Fidesz, people needed to be afraid of transgender children, immigrants and the war in Ukraine, all of which had only limited direct effects on their lives. “Disinformation and propaganda worked for many years because, most of the time, it was not made up of outright lies; there were very biased narratives based on some twisted facts. When the Orban regime lost its societal support, people became more critical of these messages,” says Hunyadi. Delegates vote during the Alternative for Germany (AfD) federal party congress in Erfurt, Germany, 04 July 2026. EPA/HANNIBAL HANSCHKE Fighting fire with fire While Fidesz used social media in what Horvath describes as a very boomer-ish way, Germany or Portugal have “young populists” in the Alternative for Germany (AfD) and Chega who can more easily mobilise on topics like house prices or young people’s emigration, unlike Fidesz after 16 years in power. However, replicating Orban’s propaganda machine won’t be easy. Political Capital’s Hunyadi argues that it needed the regime’s system, including the captured state and distorted media environment, to succeed for so long. A functioning democratic media authority and a pluralistic, competition-based media market would penalise the spreading of such disinformation, he believes. The system’s funding was also “based on corruption”. It’s telling that the massive pro-Orban media conglomerate called Mediaworks started downsizing almost immediately after the election victory of Magyar, who made it clear that his government would halt all state-backed advertising to it. Megafon, the previously heavily funded pro-Fidesz social-media influencer network, which has rejected allegations that it receives public funding, also laid off its remaining staff and closed its offices in June. “I think Hungarian society needed the experience of the last one and a half decades under the Orban regime to come up with this solution,” Hunyadi says. While Horvath doesn’t want other countries to have to experience 16 years of ‘Orbanism’ to gain this awareness, she stresses that what’s happening with the news consumption with social media, AI content and a 24/7 news cycle isn’t helping. Instead of reading a piece, processing and synthesising it, we are relying more on our emotions, she warns. “Currently, social media, especially TikTok, is extremely biased against the current order,” Katkus stresses. “[Russian disinformation narratives are] dominating German and Hungarian political content on TikTok, and we saw similar patterns in the UK, too.” As a result, narratives frequently shared on TikTok often differ from average opinions. And social media news consumption among German citizens is associated with a higher propensity to vote for populist parties, such as the far-right AfD across multiple generations and the far-left Die Linke among younger demographics. Katkus suggests “fighting fire with fire,” meaning that “democratic politicians” should prioritise short-video formats and shouldn’t rely only on press releases and stage-managed performances: “Be like Peter Magyar. He was viral and extremely good at grassroots communication.” Hunyadi highlights how Magyar’s busy in-person campaigning significantly contributed to ending the perception of the government’s invincibility. He adds that pre-bunking and pre-emptive communications were also very important, and are generally helpful against disinformation and Russian influence. The Repsense CEO praises Italy’s recent steps to anchor foreign information manipulation and interference in its national security framework, and urges smarter regulation elsewhere, too. He stresses that instead of banning individual platforms, posts or profiles, the focus should be on detecting coordination. Horvath calls for regulation against hostile propaganda and other political content using AI. She highlights that fact-checking could be better established, pointing to Portugal, where televised fact-checking is closely integrated into political debate coverage. She adds that campaign spending limits, closing loopholes in the EU’s new European Media Freedom Act, and education on the importance of democratic activities could also help. But we will never get rid of propaganda entirely. “You can’t defeat propaganda,” says Horvath, explaining that it is based on hatred, fear and a “tribal way of thinking”. While we should fight it with critical thinking, it exploits instincts that have been with us since ancient times, she warns.