The mid-twentieth-century generation experienced the boom of journalism as an effective “fourth estate” and guarantor of freedoms. Contemporary generations are witnessing its death. The fourth estate, a term which Edmund Burke allegedly coined, quickly became an essential pillar of modern democracies: information is what allows individuals to fully exercise their rights and defend themselves against the overwhelming power of the State.
But today, it seems that journalism is declining, and artificial intelligence may become the new fourth estate. But not only does AI fail to protect individuals against government power—it often reinforces the supremacy of the State. There is an enormous cultural gap between the journalistic mythomania of The Front Page, Billy Wilder’s unforgettable masterpiece, and today’s deluge of monochromatic content and news, stamped with the disclaimer “content created with the help of AI.”
Chartbeat’s latest report on audiences in the first half of 2026 reveals that search engine traffic to small digital media outlets—those receiving between 1,000 and 10,000 daily pageviews—has fallen by 60 percent over the past two years, while search engine traffic to large media outlets with more than 100,000 daily visits has dropped by 22 percent. Publishers who were counting on AI to become their new traffic provider have already been disillusioned. It is true that visits referred from platforms like ChatGPT are increasing eightfold every year, according to Similarweb data, but the reality remains bleak for mainstream media: in the specific case of ChatGPT, it represents less than 0.1 percent of the total traffic received by newspapers and media websites. Yes, AI platforms generate traffic, but they destroy far more traffic than they generate, because although they draw from the media, they offer more complete and accurate answers than conventional information, so users stop at the answer to the query and do not delve further into the sources.
This feels like the end of a long road. The idea that the press was a fundamental counterweight in a democracy vanished the day the journalistic side of the scale ceased to carry weight, because it no longer significantly influenced public opinion. The story of how we got here, while depressing, is also fascinating because it helps us understand many things about our time and how our freedoms have been partially stolen before our very eyes.
The Golden Age of Journalism, often identified with Watergate, New Journalism, and the investigative reporting of the ‘80s and ‘90s, has not been followed by a descent into a Bronze Age of journalism, but rather by an Age of Smoke. In a way, Balzac’s prophecy in Lost Illusions (1843) has come true, although for reasons different from those he could have foreseen: “Give any newspaper time enough, and it will be base, hypocritical, shameless, and treacherous; the periodical press will be the death of ideas, systems, and individuals; nay, it will flourish upon their decay. It will take the credit of all creations of the brain; the harm that it does is done anonymously.” We can identify four critical moments in this decline.
First, the rise of the Internet and digital media led to a decline in print readership. Now, any new free digital outlet could capture attention with a major exclusive, often delivered faster, more accessible, more up-to-date, and in a richer multimedia format than legacy outlets.
An army of AI-generated bots can drastically alter public opinion on social media about any hot topic in just a few hours, and can prove more effective than an extensive, well-founded exclusive report in any traditional newspaper.
Second, in the midst of this process, declining print readership and subscriptions brought a fall in advertisers, leaving many major Western newspapers in conditions of economic vulnerability; according to Pew Research, print advertising in the United States fell by more than 70 percent between 2005 and 2020. This led to the formation of large media groups encompassing newspapers, magazines, television channels, radio stations, and digital platforms, followed by a wave of layoffs and budget cuts
Third, social media burst onto the scene just as digital outlets were beginning to understand the importance of search engine optimization, and consumption habits rapidly shifted toward these platforms. It no longer mattered how a news item ranked on search engine results, but rather whether it went viral on social media, as a succession of algorithms gradually reduced the visibility of what newspapers sought to highlight, instead rewarding what users engaged with most heavily.
This degenerated into a clickbait frenzy—and with it, the old written journalism we once knew effectively died, morphing into something closer to tabloid television. In any major newspaper, a celebrity in a thong on the beach generates ten or twenty times more engagement than a major political scoop. Moreover, the thong story is ten times cheaper to produce than serious investigative reporting or war coverage.
This leaves what was once considered the powerful world of journalism in a somewhat embarrassing position. In fact, most major media outlets have removed “most-read news” rankings from their front pages to avoid them becoming a predictable mix of bulk events, celebrity news, how-to articles, sex, and emotional clickbait. Instead, they now offer “reading trends,” “most popular,” or “recommended stories,” which allow them to apply a “quality reads” filter and take other factors into account.
I am not suggesting that digitization was inherently an obstacle to quality and freedom—on the contrary, the ability to access and verify information across multiple sources has expanded exponentially. The question is whether this new landscape still allows us to consider the press a true “fourth estate.” The case is difficult to argue when, in terms of sheer influence, that role is now more often played by influencers, YouTubers, or bots deployed by governments, pressure groups, and other interested corporations to shape online debate.
The press no longer holds a monopoly on information and is frequently outpaced in both speed and scope by other channels. P. J. O’Rourke saw this early on, writing in The CEO of the Sofa: “The web is just a device by which bad ideas travel around the globe at the speed of light.”
X is a clear example of how the democratization of information can be both an opportunity and a threat to freedom. When Elon Musk acquired the platform in October 2022, he claimed his goal was to restore free speech and end moderation policies that, de facto, amplified certain ideological messages at the expense of others. X can serve as a guarantor of free expression—or as a powerful tool for manipulation by elites and governments. At different moments, it has arguably been both.
On one hand, X, Facebook, and other twenty-first-century information platforms may surpass traditional newspapers at their peak in terms of influence over immediate political cycles, daily ideological agendas, and reputational dynamics, but they have never managed to match the long-term structural influence of traditional newspapers on public opinion as a whole.
On the other hand, in 2026, we are privileged witnesses to a new paradigm shift that goes beyond social media. AI, with its amalgamation of sources and its simple, concise explanations, is rapidly gaining ground on traditional media in the role of providing information to users; and although almost all chatbots offer links to their sources, as I have pointed out previously, most users do not go beyond the first level of inquiry (i.e. the text generated directly by the chatbot) and ignore the original sources. That is to say, real influence no longer lies with media groups, nor even with the big tech companies that control search algorithms, but with AI, subject to whatever biases its developers choose to introduce at any given moment; as has already been widely debated and documented when analyzing the pro-woke biases of ChatGPT or Claude. To make matters worse, AI can explain a news story better than most journalists, and it certainly allows users to fully personalize their interests. We could be, in other words, thrown into a near future in which even the traditional digital press format could cease to make sense, or at least become something marginal.
As for print, consider the shift in the profitability model. In the era of classic journalism, ads and subscriptions were essential; today, most share a hybrid model that includes print and its advertising—the diversification of digital content, subscriptions, and digital advertising. The print newspaper is no longer the business itself, but a small product within a much broader digital business. There is a clear parallel with the music industry. Platforms like Spotify have rendered physical media largely obsolete, leaving vinyl as a niche market for romantics and collectors. Much of traditional journalism now occupies that same space.
And as if this perfect storm were not enough, we must also reckon with the rise of twenty-second news—delivered by TikTokers, tweets, influencers, or AI-generated summaries. Audiences are no longer accustomed to reading long-form analysis. Video steadily displaces text across platforms. Reality compressed into just a few seconds is far more susceptible to manipulation by individuals, organizations, or governments seeking to shape public debate.
Originally, a certain degree of transparency in funding and independence was essential for the press to function as a counterweight to power. Today, independence, both in large corporations and among small freelancers and influencers, is scarce and, in any case, very difficult to identify with any precision amid the vast media landscape. As has been noted, many of the new leading voices in journalistic opinion are flourishing on social media. They are often not journalists, have not worked their way up through the profession, and their funding is unknown. They simply went viral and were either absorbed by established media organizations or evolved their accounts into independent media channels in their own right.
Virality often appears spontaneous; it is not always so. Many seemingly random influencers are suspected of having been deliberately boosted by political actors or organizations. Some cases have come to light, sparking media scandals. For instance, Wired revealed last August that a group dedicated to promoting Democratic discourse online offered influencers up to $8,000 a month to promote the party line, under a confidentiality agreement and with certain restrictions on their content. In May 2025, the New York Times uncovered other similar networks through which Democrats are attempting to recruit new influencers to counter Donald Trump’s dominance in the social media culture war.
We cannot say that this mix of influencers, traditional media, opinions aggregated by AI chatbots, and breaking news on major social networks constitutes a “fourth estate.”
For both suspicious and legitimate influencers, what Hunter S. Thompson once denounced still holds: “The most consistent and ultimately damaging failure of political journalism in America [is that it] has its roots in the clubby/cocktail personal relationships that inevitably develop between politicians and journalists.”
It is also not easy to know the sources or the origin of information. As the profiles of what used to be a conventional journalist have become blurred, those who spontaneously turn to reporting do not understand the value of journalistic deontological principles, or are even unaware of them. When Roger Scruton wrote “An Apology for Thinking,” he was denouncing the specific case of censorship he had suffered after a decontextualized interview in the New Statesman—but, as is always the case with the British philosopher, he was also denouncing the new censors who today work on the Internet fabricating cancel campaigns:
We in Britain are entering a dangerous social condition in which the direct expression of opinions that conflict—or merely seem to conflict—with a narrow set of orthodoxies is instantly punished by a band of self-appointed vigilantes. … We are being cowed into abject conformity around a dubious set of official doctrines and told to adopt a world view that we cannot examine for fear of being publicly humiliated by the censors, this world view might lead to a new and liberated social order; or it might lead to the social and spiritual destruction of our country.
In short, the post-digital boom crisis in journalism has only exacerbated the early symptoms of decline that some authors detected decades ago. This is the case of H. L. Mencken, who drenched the profession, like everything else, in cynicism. And yet, in his later years, he glimpsed the first warning signs in the major newspapers: “In my day a reporter who took an assignment was wholly on his own until he got back to the office, and even then he was little molested until his copy was turned in at the desk; today he tends to become only a homunculus at the end of a telephone wire, and the reduction of his observations to prose is commonly farmed out to literary castrati who never leave the office, and hence never feel the wind of the world in their faces or see anything with their own eyes.”
These are old problems amplified by the new digital ecosystem. The fourth estate, at the very least, is now a decentralized, chaotic amalgam of loudspeakers, whose fragmentation empowers the State over the citizen. This fourth estate is no longer led by a small number of identifiable, more or less transparent newspapers, radio stations, and television channels adhering to certain standards of journalistic ethics. Social media often delivers information from authors we know nothing about beyond an anonymous profile on X or another platform, with the sole and controversial endorsement of a large follower count. An army of AI-generated bots can drastically alter public opinion on social media about any hot topic in just a few hours, even using fake news or emotionally manipulative rhetoric, and can prove more effective than an extensive, well-founded, exclusive report in any traditional newspaper.
The term “fourth estate” was once used because it referred to an identifiable, relatively homogeneous group that, despite ideological differences and partisan preferences, generally exercised its role as a counterweight to political power in a fairly consistent way. Today, we cannot say that this mix of influencers, traditional media, opinions aggregated by AI chatbots, and breaking news on major social networks constitutes a “fourth estate.” Rather, in its extreme fragmentation, in the mixture of truth and fake news, and in the lack of transparency regarding authorship or verifiable sources, it is political power that gains ground, vis-à-vis a dissolved and heterogeneous media landscape. Of course, at this point in the history of communication, the losers are individual freedom and the democratic health of our societies. In other words, we may now have access to vastly more information than ever before, but we are also more vulnerable to manipulation by political power.
Faced with the vast audience aggregators of influencers, newspapers obsessed with clickbait, and AI bots generating informational noise, what we once called “old journalism”—or even New Journalism in the hands of Tom Wolfe and Gay Talese—now looks set to become a refined, decorative luxury product aimed at minorities. At the very least, beyond filling the hearts of print romantics, it will serve as a reminder that freedom must be won every day.