The open web isn’t dying of natural causes. It’s being killed — deliberately, incrementally, and often by the very people who claim to love it most. That’s the uncomfortable argument gaining traction among technologists and digital rights advocates, and the evidence is getting harder to ignore.
In a recent essay that has circulated widely among developers and internet culture critics, the French publication Ouvre-boîte laid out a damning case: the slow suffocation of the open internet isn’t some inevitable market force or technological evolution. It’s a series of choices — made by corporations, by governments, and by everyday users — that are systematically dismantling the decentralized architecture that made the web transformative in the first place.
The piece doesn’t mince words. The open web, it argues, is being hollowed out by platform consolidation, algorithmic gatekeeping, and a culture that has traded agency for convenience. And the worst part? Most of us barely notice.
The Convenience Trap: How We Traded the Web for Feeds
Here’s the core tension. The original web was messy, weird, and gloriously decentralized. Anyone could publish. Anyone could link. The architecture itself was democratic — hyperlinks were votes of confidence, and the browser was your passport to an ungoverned information space. No middleman. No algorithm deciding what you should see next.
That web still technically exists. But it’s been buried.
As Ouvre-boîte frames it, the shift happened gradually. Social media platforms became the default entry point for content consumption. Instead of visiting websites, people scroll feeds. Instead of bookmarking blogs, they follow accounts. The URL — once the fundamental unit of the web — has become almost invisible to the average user. Meta, X (formerly Twitter), TikTok, and Google don’t just host content; they mediate it, deciding what surfaces and what gets suppressed through opaque ranking systems.
The result is a web that feels open but functions like a series of walled gardens. You can still build a website. Good luck getting anyone to visit it without paying for distribution or gaming an algorithm.
This isn’t a new observation. But what makes the current moment different is the acceleration. Google’s search results are increasingly dominated by AI-generated summaries that keep users on Google’s own pages, reducing click-throughs to independent sites. A Search Engine Land analysis has documented how AI Overviews are cannibalizing organic traffic, with some publishers reporting double-digit percentage declines in referral visits since the feature rolled out broadly.
Meanwhile, social platforms have been throttling outbound links for years. Meta explicitly deprioritizes posts with external URLs. X does the same. The message to publishers is clear: keep your content inside our walls, or watch it vanish.
And people comply. Of course they do. The incentive structure demands it.
I grew up in the midwest tinkering with computers in the late ’90s, building terrible GeoCities pages and thinking the internet was the most democratic thing humans had ever invented. In some ways it was. The barrier to entry was a $20-a-month dial-up connection and a willingness to learn basic HTML. That era had its own problems — discoverability was hard, design was atrocious, and half the links on any given page were broken. But the underlying principle was sound: the web belonged to everyone who used it.
Today, the web belongs to about five companies. And we handed it to them voluntarily.
The Ouvre-boîte essay identifies several mechanisms through which this transfer happened. Centralized identity systems replaced open standards. OAuth and “Sign in with Google” are convenient, but they create dependency. RSS, the protocol that once let users subscribe to any website’s updates without an intermediary, was effectively abandoned by major browsers and platforms — not because it stopped working, but because it couldn’t be monetized. Google Reader’s shutdown in 2013 wasn’t just a product decision. It was a signal.
Content management shifted too. The rise of Medium, Substack, and similar platforms gave writers easy publishing tools, but at the cost of owning their own infrastructure. Your audience belongs to the platform. Your distribution depends on the platform’s algorithm. Your archive exists at the platform’s discretion.
Even the technical standards that govern the web have become battlegrounds. Google’s dominance of the Chromium browser engine means one company effectively controls how web standards are implemented. When Google decides to deprecate a feature or push a new API, the rest of the web follows — not because it’s the best technical decision, but because Chrome’s market share makes resistance futile. Mozilla’s Firefox, the last major independent browser engine, survives largely on funding from Google itself. The irony is suffocating.
The AI Accelerant and the Enclosure of Knowledge
The emergence of generative AI has turbocharged this dynamic. Large language models are trained on the open web’s content — billions of pages of text, images, and data created by independent publishers, bloggers, journalists, and researchers. The value extraction is staggering. Companies like OpenAI, Google, and Anthropic have built multi-billion-dollar products on the back of freely available web content, then deployed those products in ways that reduce the need for anyone to visit the original sources.
It’s a kind of intellectual strip-mining. The open web provided the raw material. AI companies refined it into products. And now those products are replacing the open web as the primary interface for information retrieval.
The New York Times has reported extensively on how publishers are watching their traffic erode as AI chatbots and search summaries answer user queries without sending visitors to the underlying articles. Some publishers have begun blocking AI crawlers. Others have signed licensing deals, effectively accepting payment for content that was previously available for free — a tacit acknowledgment that the open web’s social contract has broken down.
But here’s the thing about blocking crawlers and signing deals: these are defensive maneuvers, not solutions. They treat symptoms while the underlying disease — the concentration of power over information distribution — continues to spread.
The Ouvre-boîte essay makes a critical point that often gets lost in these discussions: the problem isn’t technology itself. It’s the choices we make about how technology gets deployed and who controls it. The web’s original protocols — HTTP, HTML, DNS — are still fundamentally open. Nobody owns them. Nobody can shut them off. The infrastructure for a decentralized web still exists.
What’s missing is the will to use it.
And the incentives to rebuild it.
Consider the state of personal websites. According to data from W3Techs, WordPress still powers roughly 43% of all websites. But the growth is increasingly in managed, commercial deployments rather than independent blogs. The long tail of personal publishing — the weird, passionate, idiosyncratic corners of the web that once made it feel alive — has been shrinking for over a decade. People don’t blog anymore. They post. The distinction matters.
A blog post lives at a URL you control. A social media post lives inside someone else’s product, subject to their rules, their algorithm, their business model. When Elon Musk bought Twitter and rebranded it as X, millions of users discovered just how precarious their digital presence was. Years of posts, connections, and communities — all dependent on the decisions of a single owner. Some migrated to Bluesky or Mastodon. Most just stayed and adapted. Convenience won again.
The advertising model deserves particular scrutiny. The open web’s original sin was its failure to develop a sustainable business model for independent publishing. Banner ads gave way to programmatic advertising, which gave way to surveillance capitalism. Today, the ad-tech infrastructure that funds most “free” web content is controlled by Google and Meta, which together capture roughly half of all digital ad spending globally. Independent publishers get the scraps — and even those scraps come with strings attached, requiring compliance with ever-changing platform policies and data collection requirements.
The result is a web where the economic logic points relentlessly toward consolidation. Small publishers can’t compete for ad revenue. Subscription models work for a handful of premium brands but leave the vast middle tier stranded. And the platforms that control distribution have no incentive to change the equation — their entire business model depends on keeping users inside their walls.
So what does resistance look like?
There are pockets of genuine effort. The IndieWeb movement, which promotes personal websites, open standards, and protocols like Webmention and Micropub, has a small but dedicated following. The Fediverse — the network of interconnected servers running ActivityPub, including Mastodon and others — represents a real architectural alternative to centralized social media. Bluesky’s AT Protocol offers another model, one that separates identity from hosting and gives users more control over their algorithmic experience.
But these alternatives remain niche. The Fediverse has roughly 10-12 million registered accounts across all platforms, compared to Meta’s three billion-plus users. Bluesky has grown rapidly but still represents a fraction of the social media market. The IndieWeb community, for all its principled advocacy, is largely populated by developers and technologists who already have the skills to maintain their own infrastructure.
Scale is the problem. The open web’s tools require effort. Centralized platforms require almost none. And in a world where most people’s relationship with technology is defined by what’s easiest, not what’s most empowering, ease of use will win every time.
The Political Dimension: Regulation, or the Lack of It
Governments have been slow to recognize the structural threat. The European Union’s Digital Markets Act and Digital Services Act represent the most ambitious regulatory attempts to date, imposing interoperability requirements and transparency obligations on large platforms. But enforcement is uneven, and the laws were designed to address competition and content moderation — not the deeper architectural problem of web centralization.
In the United States, meaningful tech regulation remains stalled. The antitrust case against Google, which resulted in a landmark ruling in August 2024 finding the company had maintained an illegal monopoly in search, could eventually reshape how search works. But appeals will take years. And even a successful breakup of Google’s search monopoly wouldn’t address the broader trend of platform consolidation across the web.
There’s also a generational dimension that rarely gets discussed honestly. People under 25 have largely grown up in an app-first world. The browser is something they use for school assignments. Their primary information interfaces are TikTok, Instagram, Discord, and increasingly AI chatbots. The concept of an “open web” — of URLs, RSS feeds, and independent websites — is as foreign to them as dial-up modems. They’re not killing the open web out of malice. They simply never knew it existed.
This is perhaps the most troubling aspect of the current moment. The open web’s decline isn’t just a technical problem or an economic problem. It’s a cultural one. An entire generation is being trained to see the internet as a collection of apps rather than a shared, open medium. And once that mental model takes hold, reversing it becomes extraordinarily difficult.
The Ouvre-boîte essay ends on a note that’s less optimistic than defiant. The open web can be rebuilt, it argues, but only if people make conscious, sustained choices to support it — by maintaining personal websites, by using open protocols, by refusing to cede all their digital activity to platforms that don’t have their interests at heart. It’s an appeal to individual agency in a system that has been engineered to make individual agency feel pointless.
I think about my dogs sometimes when I think about the web. They don’t care about algorithms or platform economics. They just want to explore, sniff around, follow whatever catches their attention. The early web felt like that — an open field where curiosity was the only algorithm. What we have now is more like a series of fenced yards, each owned by someone who decides how much space you get and what you’re allowed to smell.
The fences weren’t inevitable. We built them. And if enough people care, we can tear them down.
But caring isn’t enough. It never has been. The open web needs builders, not just advocates. It needs business models that don’t depend on surveillance. It needs browser diversity. It needs protocols that make decentralization as easy as centralization. And it needs ordinary users to understand what they’ve lost — and what they still stand to lose.
That’s a tall order. But the alternative — a web controlled by a handful of corporations, mediated by AI systems trained on our collective knowledge, and designed to maximize engagement rather than understanding — is worse than anything the early web’s architects ever imagined.
The open web isn’t dead yet. But the window to save it is closing. And the people most responsible for its decline are the ones reading this article — including the person writing it.
WebProNews is an iEntry Publication