By 2027, artificial intelligence bots might create more online traffic than people, which could change how websites, search engines, and digital systems operate.
Cloudflare CEO Matthew Prince warned that automated agents are starting to browse, compare, and gather information on a scale far beyond what humans can do.
Cloudflare sees a major change in how the web is used
Speaking at SXSW, Prince said bot activity is growing so quickly that in 2027, the amount of bot traffic online will exceed the amount of human traffic that’s online.
According to TechCrunch, this increase comes from generative AI systems that now browse the web for users.
Prince explained the scale by using a shopping example. He said a person searching for a digital camera might check five websites, but an AI agent could visit a thousand times the number of sites, or about 5,000, to do the same job.
This is real traffic and real load.
Cybernews pointed out that bots can now do in seconds what a crowd of people couldn’t in a day.
Bot traffic is no longer just about search crawlers
Prince said that before generative AI, bots made up about 20% of web traffic. Search Engine Land confirmed this number and said most of it came from standard crawlers.
Cybernews added that most earlier bot traffic was from trusted crawlers such as Google, while the rest came from hackers, scammers, and all kinds of sort of miscreants.
The main change is the huge demand for data.
TechCrunch quoted Prince, who called generative AI’s need for information insatiable.
Cybernews said the AI wave is not coming in the future — it’s already here. These reports show that automated systems are now a major force, competing with humans for bandwidth, content, and infrastructure.
Even empty websites are getting hammered by bots
Cybernews shared a real-world example to show how active bots are. In an experiment, a journalist launched a new, empty .com site, but it still got 21,620 visits from 1,400 unique users.
These were not people. The visitors were bots, from engine crawlers to malicious scanners, searching for vulnerabilities, data, or exploitable features.
This experiment shows that the issue is bigger than just AI shopping assistants or search tools. The future Prince describes is not only about helpful bots. It also involves more automated pressure from scanners, scrapers, and attackers targeting websites, no matter how popular they are.
Search, publishing, and infrastructure may all need to adapt
The impact is bigger than just traffic numbers.
If bots become the main users of the web, it could change how search works and how websites make money. If AI agents start summarizing and retrieving content for users instead of sending people to web pages, publishers and marketers may need new ways to make sure their content can still be accessed, trusted, and used.
Prince also said the next stage of the internet will need new tools.
TechCrunch reported that he imagines sandboxes for AI agents, which are temporary digital workspaces created and removed as bots do tasks like planning a vacation or comparing products. He said millions of these sandboxes could be made every second, which would require big investments in data centers, servers, and uptime.
A machine-to-machine internet is coming into view
Cloudflare’s view matters because its network supports a large part of the internet.
Cloudflare’s services are used by one-fifth of all websites, so Prince has a wide perspective on changing web traffic.
His prediction is more than just a comment at SXSW; it is a warning from a company that helps manage and protect much of the online world.
If this prediction is correct, the web may soon stop being shaped mostly by people clicking through pages.
Instead, it could become a place where bots interact with each other at machine speed, pushing search engines, publishers, and website owners to rethink the purpose of the internet.
