What is the story about?
The internet you browse today may soon look very different, not because of new apps or platforms, but because of who, or rather what, is using it. According to Cloudfare CEO Matthew Prince, bots powered by artificial intelligence are on track to outnumber human users online within the next few years.
Speaking at the SXSW conference in Austin, the Cloudflare chief laid out a future where AI agents, not people, drive the majority of web traffic. It is a shift that could redefine how the internet functions, from the way websites are accessed to how data is consumed.
The rise of generative AI has fundamentally changed how information is gathered online. Instead of users visiting a handful of websites to find answers, AI-powered agents now do the heavy lifting, scanning vast portions of the web in seconds.
Prince illustrated this with a simple example. “If a human were doing a task… you might go to five websites. Your agent… will often go to 1,000 times the number of sites,” he said. That means a single query could result in thousands of site visits, dramatically increasing web traffic.
Before the AI boom, bots accounted for roughly 20 per cent of internet traffic, with much of it coming from search engine crawlers like those used by Google. Beyond that, bot activity was largely limited to spam or malicious actors.
Now, the equation is changing rapidly. “With the rise of generative AI… we suspect that, in 2027, the amount of bot traffic online will exceed the amount of human traffic,” Prince warned.
This surge is being driven by AI’s “insatiable need for data”, as models constantly scan, learn, and refine their outputs using vast amounts of online content.
Such explosive growth in bot activity comes with its own set of challenges. More traffic means more strain on servers, data centres, and the broader internet infrastructure.
Prince pointed to the Covid-19 pandemic as a reference point, when a sudden spike in streaming from platforms like YouTube and Netflix pushed parts of the internet close to their limits. But unlike that short-lived surge, AI-driven traffic is expected to grow steadily, without slowing down.
To manage this shift, new technologies may be required. One idea involves creating temporary “sandboxes” where AI agents can operate independently. “As easily as you open a new tab… you can spin up new code,” Prince explained, suggesting a future where millions of such environments could be created every second.
For companies like Cloudflare, which provide services such as content delivery networks, security tools, and DDoS protection, this evolution presents both a challenge and an opportunity. The company is already offering tools that allow businesses to manage or even block unwanted AI bot traffic.
Ultimately, Prince sees this as more than just a technical shift. “I think the thing that people don’t appreciate about AI is it’s a platform shift,” he said, comparing it to the transition from desktop to mobile computing.
If that is the case, the internet is not just getting busier, it is being reshaped. And in this new version of the web, humans may no longer be the primary drivers of traffic.
Speaking at the SXSW conference in Austin, the Cloudflare chief laid out a future where AI agents, not people, drive the majority of web traffic. It is a shift that could redefine how the internet functions, from the way websites are accessed to how data is consumed.
Why AI bots are flooding the internet
The rise of generative AI has fundamentally changed how information is gathered online. Instead of users visiting a handful of websites to find answers, AI-powered agents now do the heavy lifting, scanning vast portions of the web in seconds.
Prince illustrated this with a simple example. “If a human were doing a task… you might go to five websites. Your agent… will often go to 1,000 times the number of sites,” he said. That means a single query could result in thousands of site visits, dramatically increasing web traffic.
Before the AI boom, bots accounted for roughly 20 per cent of internet traffic, with much of it coming from search engine crawlers like those used by Google. Beyond that, bot activity was largely limited to spam or malicious actors.
Now, the equation is changing rapidly. “With the rise of generative AI… we suspect that, in 2027, the amount of bot traffic online will exceed the amount of human traffic,” Prince warned.
This surge is being driven by AI’s “insatiable need for data”, as models constantly scan, learn, and refine their outputs using vast amounts of online content.
Can the internet handle the AI surge?
Such explosive growth in bot activity comes with its own set of challenges. More traffic means more strain on servers, data centres, and the broader internet infrastructure.
Prince pointed to the Covid-19 pandemic as a reference point, when a sudden spike in streaming from platforms like YouTube and Netflix pushed parts of the internet close to their limits. But unlike that short-lived surge, AI-driven traffic is expected to grow steadily, without slowing down.
To manage this shift, new technologies may be required. One idea involves creating temporary “sandboxes” where AI agents can operate independently. “As easily as you open a new tab… you can spin up new code,” Prince explained, suggesting a future where millions of such environments could be created every second.
For companies like Cloudflare, which provide services such as content delivery networks, security tools, and DDoS protection, this evolution presents both a challenge and an opportunity. The company is already offering tools that allow businesses to manage or even block unwanted AI bot traffic.
Ultimately, Prince sees this as more than just a technical shift. “I think the thing that people don’t appreciate about AI is it’s a platform shift,” he said, comparing it to the transition from desktop to mobile computing.
If that is the case, the internet is not just getting busier, it is being reshaped. And in this new version of the web, humans may no longer be the primary drivers of traffic.
