Digital News Report 2026 — Reuters Institute
The Reuters Institute's annual report now certifies what the Sequel Observatory has been signaling for fifteen months: the overtake of social media over news websites, the collapse of organic traffic, a business model built on page views that no longer holds. We were not prophetic. We were simply paying attention.
Fifteen editions and one constant: whoever arrives first is right
The Digital News Report 2026, published on June 18 by the Reuters Institute for the Study of Journalism at the University of Oxford, is in its fifteenth edition. Approximately one hundred thousand respondents. Forty-eight markets analyzed. It is the most authoritative reference on the global digital news ecosystem.
For those who follow this Observatory, it is an interesting read. Not a surprising one. Every significant data point the DNR 2026 brings — the decline of search traffic, the rise of AI as a news source, news avoidance, the collapse of trust in media — had already been mapped, commented on, and translated into operational implications. Months in advance.
This article is simultaneously a reading of the report and a verification. A way to show concretely the value of an Observatory that does not merely disseminate data, but interprets it ahead of the mainstream.
The Italian picture: structural shift, not fluctuation
In Italy, the DNR 2026 records a deteriorating landscape on almost every fundamental indicator.
Social media is growing again for the first time after years of decline: +6 points, after a -11% accumulated from 2020 to 2024. Facebook is the main channel at 44% (+8pp), followed by Instagram at 31% (+9pp), WhatsApp at 29%, YouTube at 21%, TikTok at 11%. Print stands at 11%. Television holds, but the trajectory is one of slow decline.
The market remains dominated by television: 73% of legacy media revenues come from TV. Rai, Sky, and Mediaset together control 69% of television revenues. Streaming has surpassed 21% of consumption. Digital platforms — Meta, Google, TikTok — have captured over 85% of digital advertising. A value transfer twenty years in the making that has now stabilized as a permanent structure of the industry.
On the ownership front, the most significant news of the quarter is the 100% acquisition of GEDI — publisher of La Repubblica, HuffPost Italia, and Radio Deejay, among others — by the Greek Antenna group (Kyriakou family), completed in March 2026. The entry of foreign capital into the heart of quality Italian publishing is not a signal to read as an isolated event: it is a symptom of a sector that can no longer attract long-term domestic investment.
Italy is also a European regulatory laboratory. AGCOM has imposed fair compensation on Meta and LinkedIn for GEDI content; FIEG has filed a complaint against Google AI Overviews over the traffic decline; the AGCOM president is preparing to flag Google AI Mode to the European Commission. On the legislative front: a new AI law approved in late 2025, a code of conduct for influencers that came into force in July 2025. Regulatory context — AGCOM / FIEG, 2025-2026
The overtake: the moment that changes everything
The most important data point in the DNR 2026 is not Italian. It is global. And it is definitive.
For the first time in the history of the Digital News Report, social media and video surpass television (52%) and news publisher websites or apps (51%) as the most used source for staying informed. Adding AI as an information access channel, "third-party" sources — those that interpose themselves between the reader and the original publisher — collectively reach 56%. Among 18-24 year-olds, social media, video, and AI are already the primary source for 52% of respondents.
The meaning of this overtake is not technical. It is cultural and economic. People no longer stay informed by browsing news websites as they did in the previous decade. They find news where they already are: in feeds, in Reels, in the responses of AI assistants. The homepage of a newspaper — what was once the bar counter of digital journalism — is no longer the natural entry point.
For publishing, this changes the boundaries of competition. You no longer compete only against other newspapers. You compete against creators, against recommendation algorithms, against LLM summaries.
The wrong question: where did the traffic go?
The instinctive response from many publishers facing these figures is predictable: how do I recover the visits I have lost? That is the wrong question.
Organic search traffic — the backbone of digital publishing’s business model for the past fifteen years — has already fallen 33% globally between November 2024 and November 2025. In the United States the decline is 38%. Publishers expect a further -43% over the next three years. Traffic is not shifting: it is being halved.
This last figure deserves specific attention. Only 42% of those who use AI for news then click through to the original source. The remaining 58% stay within the assistant’s interface: they read the summary, close the tab, move on. The phenomenon has a name: zero-click. It is not new — Google has practiced it with Featured Snippets for years — but AI industrializes it at an incomparably greater scale.
The business model built on traffic — page views times CPM advertising — is already cracked. It will not return to 2019 levels. Trying to plug a failing dam with one’s fingers is not a strategy: it is a delay.
Foreign scenarios as a map of the territory
Two markets merit specific attention because they each offer, in their own way, a projection of the path Italy might follow.
United Kingdom: consolidation as structural response
The DNR 2026 photographs a British market in full reorganization. Trust in news is at 30%, down 5 points. 73% of respondents distrust news on social media. Only 4% use AI for news and only 6% trust AI assistant responses — the lowest figures globally, which says something about the quality of public debate on AI in the country.
But what matters most is what the market is doing in response to these figures. The industry responds with consolidation: the Sky-ITV merger, Axel Springer’s acquisition of the Telegraph for £575 million, the Daily Mail building a Gen Z editorial team of native creators. 18% of British respondents follow at least one news creator per week. The BBC Charter is under review, with structural implications for public broadcasting.
United States: the destination
The US is the market that looks three years ahead of Europe. Trust in media is at historic lows: 25%, -5 points, with 15% among voters of conservative orientation. A systemic proprietary capture of media is underway: Paramount passes to Skydance (Ellison family), Bari Weiss becomes editor-in-chief of CBS, Warner Bros Discovery — which owns CNN — is the subject of an acquisition negotiation. NPR and PBS are defunded by $500 million per year. Voice of America is dismantled. Bezos’s Washington Post has laid off a third of its newsroom.
On the technical front, the “Google Zero” phenomenon is more advanced than elsewhere: the decline in organic search traffic reaches 38% and in some editorial verticals exceeds 50%. This is the empirical demonstration that the traffic-based model is not cyclical: it is in structural obsolescence.
Italy, payment, and the low-trust trap
In Italy, 8% of respondents pay for online news. It was 9% in 2024. It was 6% in some previous-year surveys. For reference: in the United States 16% pay, in the United Kingdom 10%.
The question no Italian publisher wants to ask openly is this: if almost no one is willing to pay for information, what is the economic model built on?
A low-trust, low-direct-monetization ecosystem is one in which sources fragment faster. When the reader does not pay, the publisher depends on advertising. When advertising is captured 85% by platforms, the publisher depends on traffic. When traffic falls 33%, the publisher cuts quality. When quality falls, trust falls. When trust falls, even fewer readers are willing to pay. The circle closes.
This is not a prophecy. It is the description of a mechanism already underway. The way out does not pass through more content at lower cost. It passes through less content, better, aimed at an audience that considers it valuable enough to want it.
Who trusts whom in Italy: the brand rankings
In a context of systemic low trust, the distribution of reputation among individual outlets becomes a decisive competitive variable. The DNR 2026 offers precise data on this point. The reading is instructive.
| # | Outlet | Trust |
|---|---|---|
| 1 | 74% | |
| 2 | 64% | |
| 2 | 64% | |
| 4 | 59% | |
| 5 | 58% | |
| 6 | 56% | |
| 7 | 55% | |
| 8 | 54% | |
| 8 | 54% | |
| 8 | 54% | |
| 11 | 46% | |
| 12 | 44% | |
| 13 | 41% | |
| 13 | 41% |
The pattern is clear. At the top of the ranking we find the wire service (ANSA) and outlets perceived as structurally neutral or specialist (Il Sole 24 Ore, SkyTG24). Editorial polarization is penalized: the most politically inflected outlets — in either direction — slide to the bottom of the table with figures around 41-46%.
In the era of algorithms and AI, perceived neutrality and topical authority are the factors that determine whether a source is cited or ignored. It is no accident that ANSA is first: it is structurally designed to be a source of sources.
We said it already: fifteen months of lead time
The value of an Observatory is not measured in articles produced, but in the ability to anticipate trends before they go mainstream. Every significant theme in the DNR 2026 was covered in these pages, with data and operational implications, months before the report was published.
The Observatory's starting point. "Being a brand is no longer enough: you need to become a source" recognized by LLMs. In the era of AI search, positioning on search engines gives way to something harder and more durable: being the source systems cite when someone asks a question about your sector. The DNR 2026 certifies that 58% of AI sessions generate no click to the original source. All the more reason to be the source, not the article.
Italy's news avoidance at 36% — 47% among 18-24 year-olds — already flagged in this article more than a year before the DNR 2026 survey confirmed both figures. The vicious cycle described then — fewer readers = less revenue = lower quality = less trust — is today documented by Reuters data. The real enemy was not the algorithm: it was irrelevance.
Commenting on the DNR 2025, we wrote that in the US social media and video had already surpassed TV and websites as the main source of news, that only 9% of Italians paid for news, and that "journalism as we have known it will not return." The overtake the DNR 2026 now certifies globally (54% social and video) was already visible in the US a year earlier. We said it explicitly.
"Being visible is the minimum; being chosen is the achievement." SEO as a discipline of reputation, not technique. With search traffic in free fall (-33% in the following twelve months), this shift in perspective was not theoretical: it was operational.
Traffic to AI search +42.8% year-on-year (SimilarWeb/Wix, Q1 2026, over 27 billion visits). The economics of relevance in the post-ranking era: it is not whoever ranks at the top of Google who wins, but whoever is chosen by AI as an authoritative reference.
"From being indexed to being cited." Zero-click at 58-62%, AI Overview on 48% of searches. Italy's trust in news at 36% and only 6% paying — a figure updated two weeks ahead of the DNR 2026.
What remains standing: brand, audience, choice
Remove organic search traffic. Remove the display advertising model. Remove the reader who opens a browser and types in the outlet’s name. What is left?
Brands remain. Models of small, loyal, paying audiences remain. Communities of readers who choose a source not because an algorithm suggested it, but because they trust it enough to seek it out actively. It is a smaller audience, but a more stable one. And in an ecosystem where visits are being halved, stability is worth more than scale.
The data confirm it: among Italian under-35s, Il Post reaches 15% weekly reach and Will Media 11%. These are apparently modest numbers, but they correspond to audiences that have actively chosen that source — not stumbled upon it. Fanpage’s model is different: maximum online reach (22%), but low trust (41%). The choice between scale and reputation is not neutral: it defines who you will be in five years.
In the age of AI, the value of a source is no longer measured in monthly visits. It is measured in citations, in trust, in the probability that an artificial intelligence system will use it as a reference when answering a relevant question. The goal is to become a source. Not a website.
There are no universal strategies. There are sensible directions of travel in a given context. This is the context: structurally declining traffic, AI intermediating the relationship with the reader, trust that rewards neutrality and specialization, direct monetization still very low in Italy.
In this context, the concept that has guided our analyses for fifteen months is that of algorithmic reputation: the ability of a media brand — or of any organization that produces authoritative content — to be cited, not just indexed. To be the answer, not just a result. To build a presence that survives the click, because the click is becoming increasingly rare.
Frequently asked questions
