Is the internet “Dying”? Concerns Rise Over AI-Generated Content
Sam Altman, CEO of OpenAI - the company behind ChatGPT – has recently acknowledged the growing plausibility of the ”dead internet theory,” a long-held belief frequently enough relegated to the realm of conspiracy.The theory posits that a notable portion of online content is now automatically generated, effectively rendering the internet largely inactive in terms of genuine human interaction.
Altman previously dismissed the idea, but recently stated he now believes “some things may be real,” noting a considerable increase in Twitter accounts managed by Large Language Models (LLMs). This admission has fueled discussion among experts who suggest the internet may be closer to a state of “dying” than previously thought.
The first visible signs of this shift began appearing in 2024, with social media users encountering increasingly complex images created by artificial intelligence. While seemingly innocuous, these images were interpreted as indicators of AI-generated content beginning to displace human-created content. Jake Renzella of the University of New south Wales and Melbourne University, and Vlada Rozova, writing in The Conversation, noted that some of these hyper-realistic visuals garnered over 20,000 likes and comments, supporting the idea that AI-produced content is gaining prominence.
Data from cybersecurity firm Imperva supports this trend. Their 2024 report revealed that nearly half (49.6%) of all internet traffic originated from automated systems – a significant increase from 42.3% in 2021. If this trajectory continues, bots could become the majority source of internet traffic by the end of the 2020s.
Further evidence comes from the Pew Research Center,which found that 38% of web pages created by humans in 2013 are no longer accessible,a phenomenon known as “link rot.” Popular Mechanics suggests these developments reinforce the idea that the internet is evolving from being “by people, for people.”
Renzella and Rozova attribute much of this activity to “interaction farming” driven by bots, capitalizing on the equation of internet attention with revenue. Taylor Lorenz, a former reporter for The New York Times and The washington Post, described the rise of algorithmic content as the beginning of the internet’s “terminal” decline, arguing that AI-powered ranking systems have paved the way for an influx of “endless, worthless content.”
The impact extends to news dissemination. A May 2025 report by Newguard identified over a thousand news sites almost entirely managed by bots. Alarmingly,167 of these sites masquerade as Russian local news sources,spreading disinformation about the war in Ukraine using AI-generated content.
While Altman doesn’t foresee the disappearance of personal messages or shared experiences,he cautions users to be more discerning in differentiating between authentic and automated content. Renzella and Rozova are more pessimistic, suggesting the internet as we know it has “three years of life” remaining, lamenting the loss of the free exchange of ideas that once defined the platform.
