Fastly Says AI Traffic Grew 6.5x Faster Than Human Use in 2026

2 articles · Updated · TechRadar · Jun 17

AI requests on Fastly’s edge network rose about 30% in the first five months of 2026, far outpacing human traffic and signaling that AI systems are increasingly visiting websites on users’ behalf.
85% of that AI traffic came from crawlers, but Fastly said the more consequential shift is in live AI fetches—still 15% of requests—which pull fresh data for tools like ChatGPT and Gemini.
51% of AI requests hit origin servers, versus fewer than 9% of human requests, making AI traffic far less cacheable and raising hosting and infrastructure costs.
Claude-related traffic jumped 555% from January to May, reinforcing Fastly’s view that agentic AI is accelerating a broader change in how sites manage discoverability, customer acquisition and AI search visibility.
Cloudflare and Tollbit have reported similar patterns, suggesting publishers and businesses may need to monetize AI access directly as human search traffic faces structural pressure.

Sources

TechRadar1d ago

Fastly: AI Traffic Grows 6.5 Times Faster Than Human Users, Reshaping Internet

Instagram1d ago

🚨 AI NEWS ROUNDUP 🚨 The AI race is moving faster than ever. 🔹 Google's Gemini Co-Lead joins OpenAI 🔹 China's AI companies are closing the gap 🔹 The AI talent war

Fastly's leaders are selling stock while predicting an AI internet boom. What does their trading reveal about this future?

As AI replaces human clicks, how can websites survive if their content is used without providing direct traffic?

With AI agents set to dominate traffic, who will govern the new rules for digital identity and commerce online?

AI Traffic Surges 65x Faster Than Human: How Automated Interactions Are Reshaping the Web, Business, and Security in 2026

Overview

By June 2026, AI-driven traffic has become the dominant force on the internet, fundamentally transforming online interactions. This shift requires businesses and infrastructure providers to rethink their strategies, moving beyond basic bot mitigation to more sophisticated management of automated interactions. The challenge now is to identify which machine activities to accelerate, manage, or block, as AI traffic changes how the internet operates. Fastly’s reports highlight two main types of AI activity: crawlers, which systematically collect and refresh data for AI models, and fetchers, which respond to user requests. Understanding these categories is crucial for adapting to the AI-centric web.

...