Growing bot traffic and chatbot use fuel concerns about human connection online, while AI-generated disinformation poses wider risks.
The dead internet theory describes an online world increasingly dominated by bots and AI-generated content rather than people. Growing automation has given the idea fresh attention, alongside concerns about misinformation and the loss of human connection.
A 2026 report from cybersecurity group Thales found that bots accounted for 53 percent of web traffic. That figure rose from 51 percent the previous year. However, automated traffic alone does not establish the theory’s broader claim that the internet has effectively lost its human presence.
Where the dead internet theory began
The idea emerged as a conspiracy theory on internet forums in the late 2010s. It gained a clearer framework through a 2021 post on the online community Agora Road.
Supporters argued that the internet had effectively died around 2016, blaming increasingly artificial search results and social feeds. Those claims reflected a sense of emptiness online, rather than an established date when human activity disappeared.
Meanwhile, chatbot use has expanded. Pew Research found that 49 percent of Americans had used tools such as ChatGPT, Gemini or Copilot in 2026. That compares with 33 percent in 2024. The same research found that 25 percent used chatbots daily.
Automation also reaches beyond social feeds into education and recruitment. Students can generate essays with AI, while teachers use AI to grade them. Similarly, jobseekers can create applications with chatbots before automated systems assess those applications.
Disinformation risks and efforts to identify AI content
These exchanges can reduce direct human involvement, but the concerns extend beyond impersonal writing or repetitive posts. Bots can spread harmful disinformation, and AI chatbots increasingly serve as sources of false election information.
Government agencies have also warned about deepfakes and other AI-related threats to free elections. Such risks make distinguishing human communication from automated output more consequential than simply judging whether a post sounds natural.
Meanwhile, public unease has grown alongside adoption. A Pew Research study found that 40 percent of Americans expected AI to harm society over the next 20 years.
Businesses are also developing tools to identify artificial content. AI-detection company Pangram received $9 million in backing in July 2026. Following the passage of the EU AI Act, OpenAI also began watermarking AI-generated content.
