For the first time in the history of the World Wide Web, automated bot traffic has surpassed human-generated traffic in volume of HTTP requests. Cloudflare CEO and co-founder Matthew Prince confirmed the milestone on June 3, 2026, citing data from the company’s infrastructure monitoring.
According to Prince, automated traffic now accounts for 57.5 percent of all HTTP requests online, while human users generate the remaining 42.5 percent. The figures come from Cloudflare’s real-time data over the past seven days and mark a shift the company said occurred faster than industry projections had anticipated. Previous estimates had suggested this crossover would not happen until 2027.
Cloudflare attributes the surge primarily to the rapid expansion and deployment of artificial intelligence agents. These AI-powered systems browse, search, and complete multi-step tasks on the web at rates far higher than typical human users. The trend includes search crawlers and other automated agents, not only malicious bots.
Prince emphasized that the metric reflects HTTP requests, not overall time spent online or total user engagement. Humans still dominate activities such as streaming, social media, and app usage. The data shows more page requests are now triggered by machines, but people remain the main consumers of internet content.
The milestone has implications for publishers, platforms, and web infrastructure. Agentic traffic tends to generate higher volumes of requests per task, which can affect server loads, access policies, and security defenses. Companies may need to adjust how they manage bot access, rate limits, and authentication to handle the new traffic mix.
Cloudflare has also reported blocking more than 416 billion AI bot requests since July 1, 2025, according to Prince at WIRED’s Interview event. That figure underscores the scale of automated activity the company sees crossing its network.
The shift does not mean bots have replaced humans in online engagement. Rather, it marks a milestone in machine-driven web activity. The internet was not built for this volume of automated requests, and the change is expected to reshape how websites and platforms manage access going forward.
Read Article: GitHub Disables 73 Microsoft Repositories Following Credential-Stealing Miasma Campaign
