Dead Internet theory

The dead Internet theory is a conspiracy theory that asserts that, since around 2016, the Internet has consisted primarily of bot activity and automated content manipulated by algorithmic curation. This alleged coordinated effort aims to control the population and reduce genuine human interaction. Supporters of the theory claim that social bots were deliberately created to manipulate algorithms and enhance search results to influence consumers. Some proponents also accuse government agencies of using bots to shape public perception and opinions.

The dead Internet theory gained renewed interest following the AI boom that began in the 2020s, with large language model (LLM) chatbots and text-to-image models emerging as technologies that could theoretically drown out human-authored content on the web. In the time since, social media sites have seen a measured increase in bot activity, such as algorithmic feeds displaying low-quality AI slop at the expense of user-generated content.

Despite there being no evidence of a conspiracy, commentators have linked some aspects of the dead Internet theory to this rise in generative content across social media. Sources see the theory as having some amount of truth behind it, or as offering a potentially realistic prediction of the Internet's future. One source uses the term "Dead Internet" to describe spaces online that host generative content, explicitly dropping the word "theory."

Origins and spread

Academic literature struggles to capture phenomena related to online subcultures and conspiracy theories, for a variety of reasons. The dead Internet theory's exact origin is therefore difficult to pinpoint. In 2021, a post titled "Dead Internet Theory: Most Of The Internet Is Fake" was published onto the forum Agora Road's Macintosh Cafe esoteric board by a user named "IlluminatiPirate", claiming to be building on previous posts from the same board and from Wizardchan, and marking the term's spread beyond these initial imageboards. The conspiracy theory has entered public culture through widespread coverage and has been discussed on various high-profile YouTube channels. It gained more mainstream attention with a September 2021 article in The Atlantic titled Maybe You Missed It, but the Internet 'Died' Five Years Ago. This article has been widely cited by other articles on the topic.

In 2023, the dead Internet theory entered academic literature when a book published by the CRC Press included a definition of the dead Internet theory in its glossary, and in 2024, when an opinion piece titled Artificial influencers and the dead internet theory was published in the Curmudeon Corner of AI & Society. The glossary definition discussed the full theory, while the opinion piece did not, focusing instead on AI-generated content and AI-driven Interactions. These two sources have been cited by other academic articles that discuss the topic.

Claims

The dead Internet theory has two main components: that organic human activity on the web has been displaced by bots and algorithmically curated search results, and that state actors are doing this in a coordinated effort to manipulate the human population. The first part of the theory is described as the main argument, and the second where the conspiracy portion begins. This first part, that bots create much of the content on the Internet and perhaps contribute more than organic human content, has been a concern for a while, with the original post by "IlluminatiPirate" citing the article "How Much of the Internet Is Fake? Turns Out, a Lot of It, Actually" in New York magazine. The dead Internet theory goes on to include that Google, and other search engines, are censoring the Web by filtering content that is not desirable by limiting what is indexed and presented in search results. While Google may suggest that there are millions of search results for a query, the results available to a user do not reflect that. This problem is exacerbated by the phenomenon known as link rot, which is caused when content at a website becomes unavailable, and all links to it on other sites break. This has led to the theory that Google is a Potemkin village, and the searchable Web is much smaller than we are led to believe. The dead Internet theory suggests that this is part of the conspiracy to limit users to curated, and potentially artificial, content online.

The second half of the dead Internet theory builds on this observable phenomenon by proposing that the U.S. government, corporations, or other actors are intentionally limiting users to curated, and potentially artificial, AI-generated content, to manipulate the human population for a variety of reasons. In the original post, the idea that bots have displaced human content is described as the "setup", with the "thesis" of the theory itself focusing on the United States government being responsible for this, stating:

The U.S. government is engaging in an artificial intelligence-powered gaslighting of the entire world population.

A 2025 chapter in the book Market-Oriented Disinformation Research described the theory as having a "weak" and "strong" version. The "weak" version of the theory asserts that there is a group of elites using bots to shape public discourse, while the "strong" version of the theory asserts that society itself collapsed because of some catastrophic event, and some entity is keeping people connected to the internet to disguise this reality.

Expert view

Caroline Busta, founder of the media platform New Models, was quoted in a 2021 article in The Atlantic calling much of the dead Internet theory a "paranoid fantasy", even if there are legitimate criticisms involving bot traffic and the integrity of the Internet, but she said she does agree with the "overarching idea". A 2021 Ouest-France article, which heavily referenced the 2021 article in The Atlantic, went on to compare the dead Internet theory to flat Earth and COVID-19 conspiracy theories. The article stated that even though bots do produce online content, the dead Internet theory is still not realistic. In 2023 in The New Atlantis, Robert Mariani called the theory a mix between a genuine conspiracy theory and a creepypasta. A 2023 book published by CRC Press discussed the dead Internet theory, specifically mentioning Google censoring the web. The book included an entry for the term in its glossary defining it as:

The Dead Internet Theory is a conspiracy theory that suggests the Internet has died and that much of the content we see online is now artificially generated by AI to manipulate the world population. The theory raises concerns about the impact of AI on propaganda, art, and journalism.

A 2024 IFLScience article stated:

Like all good conspiracy theories, the Dead Internet Theory takes a kernel of truth or agreed sentiment (that the internet is getting worse, and that bot activity is increasing) and twists it into something it isn't.

In 2024, "dead Internet theory" was sometimes used to refer to the observable increase in content generated via large language models (LLMs) such as ChatGPT appearing in popular Internet spaces without mention of the full theory. In a 2024 opinion column in AI & Society's "Curmudgeon Corner", Yoshija Walter stated that the once speculative theory is now observable with the introduction of AI generated content. In a 2025 article by Thomas Sommerer, this portion of the dead Internet theory is explored, with Sommerer citing Walter and calling the displacement of human generated content with artificial content "an inevitable event". Sommerer states the dead Internet theory is not scientific in nature, but reflects the public perception of the Internet. Another article in the Journal of Cancer Education discussed the impact of the perception of the dead Internet theory in online cancer support forums, specifically focusing on the psychological impact on patients who find that support is coming from an LLM and not a genuine human. The article cited both Walter and the CRC Press book when defining the dead internet theory, but did not mention the conspiracy aspect. The article also discussed the possible problems in training data for LLMs that could emerge from using AI-generated content to train the LLMs. In a 2025 paper, Roland Leikauf described the dead Internet theory as "pseudoscientific" while questioning if new AI tools would justify our fear that the theory might become reality. Leikauf cites Walter's 2024 publication for his definition of the dead Internet theory. In a chapter of the 2025 book Market-Oriented Disinformation Research, it states:

What makes the Dead Internet a nameworthy conspiracy is that even though it is rooted in selective truths that are exaggerated or even taken to their logical extremes, it also draws attention to a legitimate problem.

In a 2025 interview with Time, linguist Adam Aleksic stated that the dead Internet theory "used to be a lunatic fringe conspiracy theory, but it's looking a lot more real".