R.I.P. World Wide Web, how AI is becoming a killer
AI is reshaping the web faster than any past disruption. Two dynamics stand out. On one side, platforms keep users inside their walls with zero-click results and AI overviews. On the other hand, the internet is filled with synthetic “slop” that makes trust and originality harder to sustain. Together, these shifts threaten the economic engine of the open web (+ what it means for e-commerce).
What you’ll learn in this edition
The broken bargain
The web once relied on a simple exchange: publish content, get traffic, monetise that traffic. AI is breaking the chain.
Cloudflare CEO Matthew Prince warns that generative AI and new search paradigms are killing the web’s business model by removing the need to visit websites at all. He points out that Google used to send one visitor for every two pages it crawled. Today it can take six. Large models like OpenAI’s scrape thousands of pages per prompt and send almost nothing back.
You can find more on the topic and see what people on Reddit think in the link below:
In fact, about 75% of Google queries now get answered without the user leaving Google. Google’s AI Overviews show the shift clearly. Summaries answer user intent at the top of results, often with images taken from blogs, but without a click. Pew research backs this, and Wired has shown how such features cannibalise clicks that once sustained media outlets. Social media platforms similarly discourage external clicks, favoring content hosted internally.
Slop and “enshittification”
Beyond just search, another way AI may be “killing” the internet is by flooding it with low-quality, automatically generated content, sometimes referred to as “AI slop.” This term “slop” denotes the next evolution of spam: text, images, videos generated en masse by AI with little human curation or value. Guardian columnist Arwa Mahdawi warns that “the internet is rapidly being overtaken by AI slop,” making genuine human-created content harder to find.
Cory Doctorow calls this “enshittification”: platforms optimising for clicks at the expense of quality. Nieman Journalism Lab reports that AI-driven spam accounts even gained reach thanks to Facebook’s algorithms, while MIT Tech Review warns of feedback loops where AI trains on AI-generated junk, further degrading quality.
Low-quality ‘slop’ generated by AI is crowding out genuine humans across the internet, but instead of regulating it, platforms… are positively encouraging it. Where does this end?
Arwa Mahdawi, Guardian columnist and author
The “dead internet” theory
Amid these trends, an old idea has gained new relevance: the Dead Internet Theory. This conspiracy theory posits that a majority of online activity and content is already generated by bots or AI, not people, effectively that the internet “died” sometime around 2016 and is now mostly fake. According to the theory’s proponents, government or corporate actors might be behind masses of social bots and algorithmically generated posts, seeking to manipulate public opinion or simply replace messy human discourse with something easier to control.
While the more grandiose claims are not taken seriously by experts (there’s little evidence of a centrally orchestrated plot to supplant all humans online), the observable facts about bot traffic are real.
Have you ever asked yourself what the chances are that you are interacting with an AI at any given moment on the Internet? It turns out, it’s quite high.
Tech commentator Ignacio de Gregorio Noblejas
Many internet users have indeed noticed that online reviews, forum posts, or even news articles often have a formulaic, machine-written feel. As generative AI tools proliferate, our baseline assumption may shift toward skepticism: assuming content might be AI-made unless proven human. In de Gregorio’s words, some fear “the Internet as we know it is dying.”
Enrique Dans counters that the web is not dying, but evolving. He expects reinvention rather than extinction: niche communities and investigative voices still matter, regulation or licensing may emerge, and an “agentic web” of standards-based bots and microtransactions could realign incentives while preserving diversity.
Both views highlight the same tension: how to preserve authenticity in a machine-saturated space.
Impact on websites and creators
One immediate, measurable effect of these AI disruptions is a sharp decline in web traffic for many sites. As more queries are answered by AI or kept within walled gardens, fewer clicks reach the independent web.
The pain is visible. A brief overview of some traffic/click declines:
Google initially pushed back on the idea that its AI features were harming the open web. In May 2025, Google’s VP of Search, Nick Fox, insisted “from our point of view, the web is thriving”, and Google’s Search team claimed that outbound click volume remained “relatively stable” year-over-year despite AI changes.
However, a few months later, Google seemed to contradict itself in a legal context. In a September 2025 court filing (part of an antitrust case), Google acknowledged “the open web is already in rapid decline,” arguing that breaking up Google’s ad business could “accelerate that decline”. This stark admission, highlighted by journalist Will Oremus as Google “can’t decide if the web is thriving or dying”, underscores what many publishers have been saying all along: the open web’s health is indeed deteriorating in the age of AI.
Pushback and experiments
Responses are emerging, though fragmented.
Underlying all these responses is a recognition that the open web is at a crossroads. As Prince said, this is an “existential threat” even to companies like Cloudflare: “If the internet stops existing, what’s left for Cloudflare to do?”. The broader community needs to find a balance where AI can progress and provide value without destroying the incentive to create the very content it relies on. That might mean adjusting legal definitions of fair use, developing metadata that credits sources in AI outputs, or using watermarking and detection to identify AI-made material so that human work can be distinguished (and perhaps privileged). Thus far, no one solution has emerged as a silver bullet, it will likely require a combination of approaches and ongoing adaptation.
What this means for the future of e-commerce
If AI is reshaping how people search, discover, and trust content, e-commerce will inevitably feel the impact. The stakes are high: product discovery, traffic, and conversion flows are built on the very web dynamics now under pressure.
E-commerce companies are already feeling the squeeze from lower traffic volumes as AI reshapes product search. As generative AI platforms increasingly act as product discovery engines (answering queries directly within chat interfaces) fewer consumers reach online stores through traditional search. This shift means less pre-purchase browsing on brand websites and a growing need for merchants to optimise visibility inside AI-driven ecosystems.
Three big shifts stand out:
Some experts argue this could be a forcing function: pushing e-commerce away from a race-to-the-bottom in ad spend and SEO hacks, and toward genuine product quality, stronger communities, and direct relationships with customers (via newsletters, loyalty programmes, private sales).
At the same time, new business models may emerge. Cloudflare’s pay-per-crawl vision suggests a future where AI assistants micro-pay retailers for product data each time they fetch it. Imagine your catalogue not just powering Google Shopping ads, but also fuelling AI agents that compensate you directly for every recommendation served.
In short: the death of the “old web” could be the birth of new e-commerce channels. Brands that invest in clean, structured, enriched product data and that experiment early with AI-driven discovery tools may find themselves not just surviving this shift, but thriving in it.
Closing thought
AI is undermining the open web’s vitality, but this is not its obituary. If systems emerge that credit and compensate human work, the internet will evolve rather than expire. If not, we risk a loop of machines feeding machines and a poorer web for everyone.
As The Economist put it, the open web faces an “existential threat” but its story isn’t over yet.
Head of Content & PR @ Lengow | Communication, Events & Media Relations Specialist | SaaS, B2B, Tech, eCommerce ✅
1wMy humble opinion: I believe that in times of AI-generated content, one thing matters above all: new, original, and value-adding content. In essence, GenAI means that we content creators have to produce even better content than before! Quality and relevance above all else, quite a demanding challenge btw! 😅 🙃