Big Tech platforms promised to connect us with all the knowledge in the world and ended up building walls. Google, Meta, TikTok, Apple and Amazon now hold the attention of billions of people inside closed environments designed so that no one leaves. Meanwhile, independent websites lose visitors, news outlets lose readers, and the vast diversity of sources that made the internet valuable is fading in silence, almost without anyone noticing.
This forces us to ask an uncomfortable question:
Will an open internet still exist once nobody leaves the apps?
A Garden With Walls
By: Gabriel E. Levy B.
In the technology world, a walled garden is a closed digital environment controlled by a single company, which decides what content we see, how we see it and what we can do inside it.
The term became popular in the 1990s, when America Online, better known as AOL, offered its subscribers a closed bundle of email, news, chat rooms and games, and many users believed that was the whole internet.
The open web was born as exactly the opposite. Tim Berners-Lee designed it so that any page could link to any other page without asking permission, and in 2010 he warned in Scientific American that closed social networks and apps disconnected from the web threatened that principle.
That same year, Chris Anderson and Michael Wolff published an article in Wired titled The Web Is Dead, arguing that apps and platforms were replacing the browser as the gateway to the internet. Many dismissed it as an exaggeration. Sixteen years later, the numbers prove them right.
The Numbers Behind the Enclosure
According to Sensor Tower’s State of Mobile 2026 report, people spend less than 6% of their phone time in browsers and search apps. The remaining 94% happens inside apps, and social media alone accumulated close to 2.5 trillion hours in 2025.
Even the search engine, which for two decades served as the great distributor of visits to the rest of the web, has started closing its doors. SparkToro calculated that 68% of Google searches in the United States ended without a single click to any website between January and April 2026.
AI-generated summaries, which Google calls AI Overviews, accelerated that process. The Pew Research Center found that only 8% of searches showing an AI summary produced a click, compared with 15% of searches that did not display one.
Matthew Prince, CEO of Cloudflare, the company that protects a large share of global internet traffic, summed it up with a brutal figure. Ten years ago Google sent one visitor to each site for every two pages it crawled; today the ratio is 18 to 1, at OpenAI it reaches 1,500 to 1, and at Anthropic it exceeds 60,000 pages read for every visitor sent back.
The News Sites Nobody Visits
The Digital News Report 2026 from the Reuters Institute at the University of Oxford, based on nearly one hundred thousand surveys in 48 countries, recorded a historic milestone. For the first time, social media and video platforms overtook news outlets’ own websites and apps as the gateway to news, 54% versus 51%.
Some 12% of respondents get their news only through social networks, double the figure from 2020, and 77% watch news videos every week, almost always on YouTube, TikTok, Instagram or Facebook and less and less on the outlets’ own pages.
The blow to the cash register is already being felt. Chartbeat measured that visits sent by Google Search to publishers fell 34% between December 2024 and December 2025, and small sites lost up to 60% of those visits.
Nicolas Bouliane, creator of AllAboutBerlin, an independent guide for immigrants in Germany, reported in May 2026 that his site had lost 70% of its traffic to AI summaries after seven years of growth. His story repeats across thousands of blogs, forums, wikis and specialized portals that nobody counts.
The Algorithm That Decides for Us
Inside these gardens, a hyper-personalized algorithm decides what appears on the screen. The platforms present it as relevance. In practice it measures retention, that is, how long we stay inside the same app.
That logic produces a curious effect. We have never had so much information available and yet more and more people feel exhausted by it.
The Reuters Institute calls this phenomenon news avoidance and puts it at 42% globally. In Colombia, nearly five out of ten people say they avoid the news sometimes or often, while worldwide trust in journalism fell to 37%, its lowest level in a decade.
In Latin America the enclosure runs even deeper. In Colombia, 35% of people get their news weekly from content creators and influencers, 9% already use AI chatbots to follow the news, and in Peru more than 80% consume news video on social networks every week.
I personally believe that information fatigue is a problem of diet. We eat what the algorithm serves, in the order the algorithm chooses, and we end up stuffed with more of the same.
The Diversity That Is Fading
When readers stop visiting pages, small sites stop existing. A gardening blog, a car repair forum, a regional magazine or a community encyclopedia survives thanks to the click that arrives from a search engine, and that click is going extinct.
The result is a more homogeneous internet, where five companies filter what billions of people see, and where artificial intelligence summarizes content that, over time, nobody will have any incentive to keep producing.
Something is moving, however. Cloudflare now blocks AI crawlers by default and is testing a pay-per-crawl system; Penske Media, publisher of Rolling Stone and Variety, took Google to court over AI Overviews; the UK competition authority is forcing Google to give publishers a say in how their content appears in AI answers; and several countries are debating rules against algorithmic manipulation.
We have part of the homework too. Subscribing to newsletters, going back to RSS readers, typing a news outlet’s address directly into the browser and paying for the journalism we value are small gestures that keep the garden gates open.
In short, the big platforms turned the internet into a collection of walled gardens where algorithms hold our attention and AI summaries answer without anyone going out to find the source. Zero-click searches, the collapse in news site traffic, the disappearance of small sites and reader exhaustion all point in the same direction. The open web is not dead yet, but it will survive only if we walk back through its gates.
References
- Anderson, C., & Wolff, M. (2010). The Web Is Dead. Long Live the Internet. Wired. https://www.wired.com/2010/08/ff-webrip/
- Berners-Lee, T. (2010). Long Live the Web: A Call for Continued Open Standards and Neutrality. Scientific American, 303(6), 80-85.
- Reuters Institute for the Study of Journalism. (2026). Digital News Report 2026. University of Oxford. https://reutersinstitute.politics.ox.ac.uk/digital-news-report/2026
- Reuters Institute for the Study of Journalism. (2026). Digital News Report 2026: Colombia. University of Oxford. https://reutersinstitute.politics.ox.ac.uk/digital-news-report/2026/colombia
- Nieman Lab. (2026, June 15). Facebook for news is on the rebound, impartial news isn’t dead, and other findings from RISJ’s 2026 Digital News Report. https://www.niemanlab.org/2026/06/news-sites-are-the-new-newspapers-people-are-abandoning-them-for-social-media/
- Pew Research Center. (2025). Google users are less likely to click on links when an AI summary appears in the results. https://www.pewresearch.org/
- Similarweb. (2026). Zero-Click Marketing: What the 2026 Data Means. https://aisearch.similarweb.com/blog/zero-click-marketing/
- Sensor Tower. (2026). State of Mobile 2026. https://sensortower.com/
- Fischer, S. (2026, March 17). Publisher traffic from Google Search fell 34% in a year, Chartbeat data shows. Axios.
- The Decoder. (2025). Cloudflare CEO Matthew Prince sees trouble ahead for the open web. https://the-decoder.com/cloudflare-ceo-matthew-prince-sees-trouble-ahead-for-the-open-web/
- Engadget. (2025). Cloudflare CEO says people aren’t checking AI chatbots’ source links. https://engadget.com/ai/cloudflare-ceo-says-people-arent-checking-ai-chatbots-source-links-120016921.html
- Penske Media Corporation v. Google LLC. (2025). Complaint filed in the United States District Court for the District of Columbia.
