Three days after securing a landslide victory in Hungary’s parliamentary election, incoming prime minister Peter Magyar appeared on the country’s state broadcaster for the first time in 18 months and labeled it a “factory of lies” peddling “propaganda” worthy of North Korea and Goebbels.

Magyar’s hostility reflects the well-documented media capture that outgoing prime minister Viktor Orbán and his Fidesz party entrenched for 16 years after returning to power in 2010. Through a combination of restrictive laws, partisan enforcement, and hostile takeovers, some 80% of Hungary’s traditional media was Orbán-aligned, hoovering up an estimated 90% of state advertising revenue. In addition, the government passed laws meant to shut down or intimidate academic institutions and civil society organizations shining a critical light on the illiberal and kleptocratic tendencies of Orbán’s Hungary.

So how did Magyar and his Tisza party manage to create a political movement winning a two-thirds majority in Parliament with a media environment so heavily stacked against opposition voices? Well, it’s the internet, Stupid! Although Magyar had a very strong ground game, visiting several villages every single day, the internet was a huge factor in his campaign.

With all print and broadcast media serving as Fidesz mouthpieces, Hungarian journalists, civil society organizations, and opposition politicians heavily relied on the internet. Small and nimble online media outlets persisted with independent reporting, documenting government scandals, using social media to provide an alternative news diet to ordinary Hungarians used to being force-fed government propaganda.

A pivotal moment happened on Feb. 2, 2024, when 444.hu broke the story of how Hungary’s president, Katalin Novák, had pardoned a person convicted for covering up child-sex abuse at a government-run children’s foster home. Investigative media outlet Direkt36 and Telex.hu dug deeper and exposed how a prominent religious leader and “spiritual mentor” to Orbán had played a prominent role in securing the pardon.

Despite silence or minimization in traditional media, the story exploded online and led to massive public protests. The “pardon affair” forced Novák to resign, followed by the resignation from political life of Judit Varga, who had been minister of justice when the pardon was granted.

The scandal became a watershed moment for Magyar’s meteoric political career. On the very day that Novák and Varga resigned, Magyar—a long-time Fidesz member, political official, and Varga’s ex-husband—gave an explosive interviewwith Partizan, a YouTube channel hosting political debate and commentary. Magyar accused Fidesz of corruption and throwing the two disgraced women under the bus to protect the real culprits. The interview quickly reached 1 million views (now at 2.8 million in a country of 10 million). Subsequent Facebook and YouTube posts by Magyar with secret recordings of Varga allegedly describing government interference in a corruption prosecution also went viral. In April, he launched Tisza as a new political party, and two years later he humiliated Orbán in a landslide election.

Social media was also central to Tisza’s election campaign. Despite being vastly outspent by Fidesz—which spent over 85% of all political advertising money on Meta platforms before the company’s October 2025 ban on political advertising took effect, and then created mass private Facebook groups to circumvent the ban—Tisza consistently outperformed Fidesz where it mattered most: organic engagement. According to ResFutura, a Polish data analytics outlet that tracked political activity on Facebook throughout the campaign, Tisza’s per-post engagement rate was roughly three times higher than Fidesz’s—meaning Magyar’s content resonated far more effectively with actual audiences, even as Fidesz posted at far greater volume and with far greater resources.

Tisza’s stunning victory challenges a decade of techno-pessimist warnings about how the online ecosystem of alternative and social media threaten democracy and necessitate more government regulation of speech and platforms. In fact, doubling down on free speech as both a legal and cultural foundation for democracy may well be a stronger antidote to creeping authoritarianism.

Again Hungary is case in point. In the 2025 Future of Free Speech Index, a 33-country survey by the Future of Free Speech think tank at Vanderbilt, where I am Executive Director, Hungary ranked third in popular support for free speech—behind only Norway and Denmark, and ahead of the United States. Yet Hungarians also reported among the sharpest declines in their actual ability to speak freely, trailing only Turks and Venezuelans. Sixteen years of Fidesz restrictions had sharpened rather than eroded the demand for what Orbán was narrowing.

Tisza’s victory highlights a truth that is obscure from the vantage point of liberal democracies but existential from the point of view of authoritarians such as Xi Jinping and Vladimir Putin: control and censorship of legacy media is a necessary but insufficient condition to entrench an authoritarian regime in the digital age.

Autocrats and Walls

Following the Arab Spring that felled dictators in Egypt, Tunisia, and Libya, both Xi and Putin became obsessively focused on countering the disruptive influence and mobilizing potential of online dissent through systematic and pervasive online censorship. Consequently, China expanded its “Great Firewall” while Russia constructed its “Red Web,” including tight and real-time government control of online discourse and blanket bans of the American tech platforms that Arab protestors had used to help topple their rulers. Both Xi and Putin would thus have regarded Orbán’s media capture and censorship regime as a half measure with a fatal flaw given its breathing room for independent news and organic mobilization via online and social media.

What liberal democratic establishment thinking gets right is that the internet and social media tend to favor ideas and persons that challenge the prevailing narratives among elite institutions in politics, legacy media, and academia. In democracies run by broadly centrist governments, that dynamic provides an opportunity for populists, very often illiberal, to gain traction with narratives that resonate with large swaths of the population but alarm elite opinion (think immigration, globalization, and identity politics).

That explains why politicians vulnerable to populist backlash—including Democrats in the United States and Emmanuel Macron, Friedrich Merz, and Ursula von der Leyen in Europe—are so invested in countering the perceived threat to democracy from online “disinformation” and the platforms that disseminate it. Disinformation by bad-faith actors who explicitly seek to “flood the zone with shit” to overwhelm our capacity of sifting truth from lies is a genuine problem. But efforts to curb it can often morph into curbing uncomfortable dissent. In the past six months alone, Merz has called for ending online anonymity, Macron has demanded new legislation to block false information online, while the EU has administratively banned former Swiss army colonel Jacques Baud from entry and frozen his assets for being “a mouthpiece for pro-Russian propaganda” and advancing “conspiracy theories.”

But dissent is the price of governing in open democracies where political power derives from the consent of the governed and where winning elections does not provide politicians with the power to determine truth or censor disfavored viewpoints.

Establishment opinion should also take heart from Orbán’s defeat, since it demonstrates that once in power, populists too are vulnerable to the backlash effects of online discontent. In America, many of the contrarian voices that boosted the MAGA “vibeshift”—from heterodox podcaster Joe Rogan to conservative commentator Tucker Carlson—have used their online platforms to criticize Trump. So have progressive online voices such as MeidasTouch and Brian Tyler Cohen, gaining massive followings with younger audiences whose news consumption skews heavily towards social media.

None of this means liberal democracies can afford complacency about their information environment. Bad-faith actors—grifters, demagogues, and state-sponsored disinformation operations—exploit the openness of free societies precisely because they do not share liberal society’s commitment to truth. This was true in Ancient Athens, when demagogues swayed the assembly into disaster; it was true when Bismarck manipulated a telegram and a gullible press to engineer the Franco-Prussian War; and it was true when The New York Times’ Walter Duranty won a Pulitzer Prize for reporting that whitewashed the Holodomor, Stalin’s mass starvation policy in Ukraine.

No mechanism catches every lie, and coverage of harmful content will always be imperfect. But there are non-restrictive strategies that leverage rather than restrict free speech that democracies can implement to mitigate the harms of disinformation.

Taiwan, which absorbs more Chinese disinformation than any European democracy faces from Russia, rejected the path of censorship. Its approach—pioneered by former Digital Minister Audrey Tang—is grounded in radical transparency and civic, crowdsourced rebuttal. Ordinary users flag suspected disinformation for real-time community debunking rather than state censorship. That Taiwanese model has since been exported and helped inspire Twitter’s 2021 “Birdwatch” pilot, later renamed Community Notes under Elon Musk, which allows users to flag potentially misleading content, resulting in a visible Community Note if a critical mass of people with diverse political views rate it as helpful.

This approach has already proved helpful. When Trump administration officials tried to reshape the narrative around the January 2026 killing of ICU nurse Alex Pretti by Border Patrol agents in Minneapolis, Community Notes allowed the American people to fact-check their government in real time, and the administration’s falsehoods and distortions quickly collapsed. This is not a mere anecdotal example. Peer-reviewed research has repeatedly shown the potential of crowd-sourced fact-checking when it comes to issues such as COVID misinformation, political misinformation, and reducing the virality of false information.

The mechanism strives to be politically neutral. It is true that under Musk, X has taken a right turn. He uses the platform to support “Great Replacement” conspiracy theories, massively boost his own views while selectively silencing opposing ones, all the while branding himself a “free speech absolutist” while behaving as anything but. A 2025 study shows that posts by Republicans are more than twice as likely to be flagged as misleading than posts by Democrats. The mechanism doesn’t work quite as well for Musk himself. He has been repeatedly community-noted on his own platform, which is great. But his dedicated fanbase turns out in droves to defend him, often managing to get the note removed, as a Bloomberg investigation found. This demonstrates that the success of crowdsourcing ultimately depends on safeguards that prevent platform owners from gaming the system.

Pro-Social Media

So, for all its promise, crowdsourced fact-checking is no panacea. It cannot cure every form of disinformation, propaganda, or conspiracy theory, and correction after the fact is only one tool for improving the online information ecosystem. As Tang, Glen Weyl, and I have argued, the more ambitious project is to redesign social media itself around “prosocial” principles—surfacing content that bridges communities rather than inflames them, giving users tools to see who is actually engaging with what they read, and making shared understanding rather than outrage the engine of engagement. As we argue:

[A] prosocial media ecosystem will encourage regulatory frameworks that align with democratic values—openness, transparency and free expression—rather than the current trajectory, where democracies find themselves on the defensive, resorting to measures long championed by closed societies to safeguard their systems of governance.

The lesson from Budapest is not that the internet and social media are inherently good or will always lead to the victory of democracy over authoritarianism. It is that subjecting them to government control is a necessary precondition for authoritarians to stay in power, since they provide a crucial outlet to circumvent official propaganda and censorship. Moreover, an overly critical tendency to characterize the online sphere as a toxic swamp irredeemably contaminated by lies and hatred ignores the huge potential of digital technology to improve our currently imperfect ecosystem of ideas and information.

Free speech is not a luxury good that democracies may ration when the information environment becomes uncomfortable. It is the precondition for the self-government those democracies claim to embody—the mechanism by which voters identify what their rulers are concealing, by which excluded constituencies signal the concerns elite consensus has missed, and by which an insurgent movement armed with a YouTube channel and a Facebook page can humiliate a 16-year authoritarian incumbent. The task for establishment institutions is not to paternalistically manage the digital public sphere but to listen to what it is telling them. Democratic governments that respond by expanding their power to silence the messengers may well find, as Orbán did, that the messages keep arriving, that the audiences keep growing, and that at the ballot box the voters remember who tried to shut them up.