For the first time in the history of the Web, traffic generated by automated software has exceeded that produced by humans: almost 6 out of 10 connections are generated by machines. This is shown by the latest data from Cloudflare Radar, according to which bots, crawlers and AI agents now represent 57.4% of web requests, compared to 42.6% attributable to human users.
To better understand what this data means, let’s imagine that a site receives 100 requests: today around 57 would come from automatic systems and only 43 from “real” people. This is a significant change, which reflects the growing diffusion of artificial intelligence-based tools capable of autonomously navigating online.
Bots, crawlers and AI agents: what are the differences?
The automated systems that generate web traffic do not all perform the same functions. Some are used to index online content, others to collect data or monitor digital services, while the latest AI-based agents are capable of performing more complex tasks on behalf of users.
Crawlers are programs used by Google and all other search engines to continuously explore the web and update their archives. When a new page appears in search results, it is because a crawler first found it, analyzed it and inserted it into the search engine’s index.
There are also numerous automatic bots used for activities such as monitoring prices, collecting public data, checking product availability or checking the functioning of websites.
In recent years, a third category has been added, destined to have an ever-increasing impact: that of AI agents. Unlike traditional crawlers, these tools don’t just visit a page, but can understand its content, compare it with other sources and perform complex tasks, such as searching for information, comparing products or collecting data needed to answer a question.
Why is auto traffic increasing so rapidly
The growth of automated traffic is closely linked to the evolution of AI-based tools. Unlike traditional crawlers, which are designed to perform specific tasks such as indexing web pages, the latest AI agents can perform complex tasks that require consulting numerous online sources. In fact, LLMs need to “eat” an enormous amount of data: so hundreds of thousands of explorer bots relentlessly scour every single site to read articles, download images and learn our behaviors.
To better understand the phenomenon, an example may be useful. If a person wants to buy a new smartphone, he will probably visit some specialized sites, compare prices and read reviews before making a decision. An AI agent tasked with the same task can instead consult many more pages in extremely short times, automatically collecting and organizing the necessary information.
This means that a single request made through an AI-powered assistant can result in numerous interactions with online sites and services. It is also for this reason that the weight of automated systems in web traffic continues to grow.
A change that came sooner than expected
The overcoming of human trafficking by automated systems is not only a symbolic achievement, but also a phenomenon that has occurred much more quickly than expected. As Matthew Prince, co-founder and CEO of Cloudflare, explained, the overtaking was initially expected for the end of 2027 and subsequently anticipated at the beginning of the same year, but the growth of AI-based agents has significantly accelerated this trend.
According to Prince, one of the main reasons is related to the different way in which humans and AI systems interact with the Web. A person visits on average about five sites before completing an online purchase, while a chatbot can compare the same product on thousands of pages in a much shorter time.
This difference helps explain why traffic generated by bots and intelligent agents is increasing so rapidly. Each request made through an AI assistant can in fact translate into numerous interactions with sites and online services, contributing to increasing the overall volume of automated traffic.
What could change for the future of the Web
The overtaking of automated systems over human users is not just a statistical curiosity. More and more often, in fact, access to online information is through tools capable of searching for content, comparing sources and carrying out activities independently.
This scenario partly recalls the so-called “Dead Internet Theory”, according to which in the future a growing share of online activities could be generated by bots interacting with each other. This is a controversial theory and often associated with extreme interpretations, but the data recorded by Cloudflare Radar shows how the role of automated systems within the digital ecosystem is actually becoming increasingly relevant.
It is still early to understand what the impact of this transformation will be in the long term. However, the growth of AI agents suggests that an ever-increasing portion of the tasks currently performed directly by users could be entrusted to software capable of operating autonomously on the Web.
