Here’s what’s on our plate today:

• 🧠 Is generative AI quietly killing the open web?

• 📌 Apple’s new iPhone Air, Microsoft’s $1.74B AI bet, and Google’s Veo 3 update.

• 🛠 Snack for Builders: Building user trust when the open web is collapsing.

• 📊 What do you think: Is the internet still alive?

Let’s dive in. No floaties needed…

The Laboratory

Is generative AI killing the open web?

The modern world is made possible through advancements in technology, many of which rest on the rapid flow of information. While it makes cooperation between distinct entities possible, the system relies on information with humans in the loop. Naturally, when a new technology like generative artificial intelligence, capable of generating text, images, and videos, was added to the existing system, it raised concerns around the viability and maintenance of the credibility of these systems. In this backdrop, when OpenAI CEO Sam Altman said he suspects that a large chunk of X’s accounts were being run by LLMs, it started an online debate.

While on the surface, Altman merely looked to be highlighting the increasing number of bot-run accounts on the social media network, the conversation has quickly shifted to questions around the impact of generative AI on the internet. This has once again brought the dead internet theory to the forefront.

Reports suggest that the search giant Google admitted in court filings that the open web is in rapid decline. The statement was part of a filing submitted by Google ahead of a trial to determine if it has a monopoly in the ad tech business. At this trial, Google, in a bid to defend itself, is arguing that breaking up its ad business isn’t ideal because it would only accelerate the decline of the open web, harming publishers who currently rely on open-web display advertising revenue. This is not the first time the Dead Internet Theory has surfaced in online forums, and it likely won’t be the last.

What the Dead Internet Theory says

The Dead Internet Theory (DIT) is the idea that the public internet is increasingly synthetic. The theory postulates that engagement is juiced by bots, content is mass‑produced by algorithms/LLMs, and the human signal is drowned out.

The concept bubbled up in 2021, when The Atlantic chronicled a wave of near‑identical social posts and the growing sense that online spaces felt automated and ‘samey’. It’s been resurfacing ever since, especially as generative AI scaled.

The Guardian reported that one proponent of the theory suggested that the internet died in 2016 or early 2017, and that now it is empty and devoid of people. While the claim remains unproven, the feeling that there are more bots on the internet than actual people sharing information is supported by measurable trends.

Bots, AI, and the numbers behind the theory

The core idea, propelling the belief that the internet has more bots than humans as active users, is one supported by numbers.

According to Imperva’s 2024 Bad Bot Report, 49.6% of global internet traffic in 2023 was automated, with nearly a third comprising bad bots.

Akamai’s 2024 report likewise pegged bots at 42% of internet traffic. Cloudflare’s application‑security update says about a third of the traffic it sees is automated, and most is not from verified (good) bots. None of this means that most people online are fake, but it does mean a massive slice of online requests is coming from machines, not humans.

And it is not just bots. AI-generated content is also quickly taking up a major chunk of the information available on the internet.

According to NewsGuard, up until May 2025, it had identified 1,271 Unreliable AI-generated news and information websites spanning 16 languages. These websites reportedly had generic names resembling those of established news sites and operated with little to no human oversight, publishing articles largely or entirely written by bots.

The sites were also found to be peddling false claims, including about political leaders and celebrity death hoaxes, fabricated events, and articles presenting old events as if they just occurred.

The goal: to generate revenue through the ad-tech industry, which often disregards the nature and quality of the content and relies heavily on site visits to compensate the publisher. Beyond the malicious use of bots and AI-generated content, AI slop, overviews, and chatbots have also shifted the focus away from credible websites, which in turn has made reliable information on the internet difficult to find.

AI slop, bot traffic, and the search for answers

One of the public-facing access points of the internet, Google Search works by crawling web pages, indexing their content, and ranking them by relevance, authority, and user experience before serving results. However, the company has struggled with AI-generated content that is written solely for the purpose of ensuring a website ranks high on Google’s search results. Content on such websites is often stuffed with keywords in hopes that Google’s algorithms will pick them up and rank the website high on the search results.

In March 2024, Google rolled out updates and new spam policies targeting scaled content abuse, site‑reputation abuse, and expired‑domain abuse. These were aimed at cutting “low‑quality, unoriginal content” in results by 40%. They hoped to reduce the AI-generated content from ranking high on its search engine, and underline the problem stemming from the misuse of generative AI tools.

All this has resulted in a decline in the reliability of the information that flows through the internet. According to a survey of 2,000 Americans, conducted by Talker Research in 2025, three-quarters say they trust the internet less today than ever before.

Among those polled, only 31% said they are confident in their ability to differentiate if a product or service review was written by an AI or an actual human being. Even when they believed that 50% of the news stories and articles they come across online have some element of AI, whether it be images or the actual written content.

Another factor in the decline of human users is the increasing use of AI-generated summaries by companies to push their products, even when they are unreliable.

Google’s AI Overviews drew ridicule for confidently giving incorrect advice (e.g., “glue on pizza”), illustrating how generative systems can amplify nonsense from weak sources. Google has since throttled and tuned the feature, but the episode sharpened concerns about information quality.

Things have taken such a turn that websites are now blocking AI crawlers from accessing information that drives traffic away from their domains. In mid‑2025, Cloudflare began blocking known AI crawlers by default and launched Pay‑Per‑Crawl so publishers can meter or monetize AI access.

The measures highlight the inherent shift in how information on the internet no longer flows through reliable sources, and users are increasingly turning to AI overviews and chatbots for information, which further reduces active human participation online

The open web’s survival question

If major news outlets are to be believed, they are facing an existential traffic hit as AI answers gain favor. Many have responded by putting their content behind paywalls, which further discourages users. However, faced with changing user behavior, they may have no other viable alternative. One way is to sign deals with AI companies and seek compensation for the use of their data; however, even this does not help with revitalizing the internet.

The dead internet theory claims that the internet is mostly used and populated by bots, and content is generated by AI. While the theory fails to account for the large number of human users who continuously generate content and interact with it, it does highlight a growing trend. Roughly 40–50% of traffic on the public web is automated.

Altman’s remark then hits the nail on the head: the internet may not be dead, but it surely is undergoing a tremendous change thanks to generative AI tools and chatbots. Where it goes from here will depend on user preferences and how search engines reinvent themselves to retain users.

On the broader front, whether the open web, as we know it, can survive in an age where information is increasingly mediated by AI is a question that will depend on how publishers reshape their economic model. So far, it has formed the backbone of the model that allowed free access to quality information, but if it collapses, it risks creating an ecosystem where only those who can afford premium subscriptions or negotiate licensing deals with AI companies have access to credible knowledge.

Meanwhile, the majority may rely on AI outputs of uncertain accuracy. In other words, AI isn’t killing the open internet outright; it is quietly reshaping it into a more closed, fragmented, and commercialized space where information no longer flows freely through credible channels, but rather depends on large corporations that may not have users’ or the internet’s best interests at heart.

Brain Snack (for Builders)

Thursday Poll

Quick Bits, No Fluff

• Apple unveils slim iPhone Air: Apple has announced several new products, including its new slimmer iPhone “Air” model with a “high-density battery”.

• Microsoft Bets $1.74B on AI Compute: Nebius AI’s stock jumped after announcing a multibillion-dollar deal with Microsoft to boost global AI infrastructure.

• Google’s Veo 3 Gets 1080p & Vertical Video: The latest version of Google’s video-generating AI supports 1080p output and vertical formats — a nod to TikTok and Shorts creators.

Meme Of The Day

Rate This Edition