Your daily web browsing just became a minority activity. Imperva’s latest Bad Bot Report reveals that bots generated 51% of all internet traffic in 2024—the first time in a decade that machines have outnumbered humans online. Before you start picturing a dystopian robot uprising, the reality is more nuanced and arguably more fascinating.
When Shopping Agents Hit 5,000 Sites in Seconds #
AI-powered browsing multiplies web requests at unprecedented scale.
The surge isn’t just about traditional crawlers scraping websites. “ Agentic traffic“—
AI systemsthat browse like humans but at machine speed—is exploding. Picture this: you might manually check five
camera retailersbefore buying. Meanwhile, an AI shopping agent working on someone else’s behalf hits 5,000 pages, comparing specs, prices, and reviews in seconds.
That multiplicative effect explains why Cloudflare’s radar now shows bot traffic consistently in the mid-50s, sometimes spiking to 62% during peak periods.
The Geography of Fake Traffic #
Bot concentration reveals the infrastructure behind automated browsing.
The numbers get wild when you zoom into specific regions. Gibraltar reportedly sees over
90% of its HTTP requests from bots, while
Singapore and
Iran clock in above
75%. These aren’t nations of robot citizens—they’re data center hubs where VPNs, proxies, and hosting services make automated traffic appear to originate.
It’s like discovering that most “New York” Instagram posts actually come from a server farm in New Jersey.
Your Analytics Are Lying to You #
Bot-heavy traffic warps business metrics and ad spending.
This affects your actual internet experience in concrete ways. Publishers struggling with bot-inflated pageviews can’t distinguish real engagement from automated noise. Advertisers pay for impressions that no human ever sees.
Imperva found that malicious bots—the truly problematic ones—account for 37% of all web traffic, with
44% of sophisticated attacks now targeting APIs that power modern websites and mobile apps.
The Feedback Loop Problem #
AI systems increasingly cite other AI-generated content, creating information distortion.
The most unsettling trend? More than 10% of AI summaries now reference AI-generated content rather than human-created sources. We’re entering a hall of mirrors where machines summarize articles written by other machines, potentially amplifying errors and biases with each iteration.
Your search results might soon be dominated by bots talking to bots about what other bots already said.
This doesn’t mean humans have abandoned the internet—you’re still binge-watching Netflix, scrolling TikTok, and buying stuff online. But the infrastructure layer increasingly runs on autopilot, with AI agents handling the heavy lifting of information gathering while you focus on the fun parts.