AI Bot Traffic by Country: Where AI Crawlers Are Most Aggressive (2026 Insights)

AI crawlers account for 4.2% of all HTML requests globally, but traffic patterns vary dramatically by country.

Published February 8, 2026 Updated June 29, 2026

AI crawlers now account for roughly 4.2% of all HTML requests hitting websites globally, according to Cloudflare Radar data. But that single number hides a dramatic geographic divide. I analyzed 30 days of Cloudflare Radar AI bot traffic data across 11 countries and found that the dominant crawler, its purpose, and its intensity vary wildly depending on where your website's audience sits. In Western markets, Googlebot commands nearly half of all AI-related traffic. In Brazil, Japan, and Australia, ChatGPT-User is the runaway leader -- and the crawling purpose shifts from model training to real-time user queries.

What Does the Global AI Bot Landscape Look Like in February 2026?

Globally, Googlebot leads all AI-related crawler traffic with a 38.7% share, according to Cloudflare Radar AI Insights (get_ai_data endpoint). GPTBot follows at 12.8%, Meta-ExternalAgent at 11.6%, and ClaudeBot at 11.4%. These four crawlers alone represent 74.3% of all identified AI bot requests hitting websites protected by Cloudflare's network of 330+ cities in 125+ countries.

One significant month-over-month shift stands out: Meta-ExternalAgent surged from 8.5% to 11.6% of global AI bot traffic between December 2025 and January 2026 -- a 36% relative increase in just 30 days. Meanwhile, Googlebot's share declined from 42.6% to 38.7%, suggesting Meta is ramping up its training data collection aggressively while Google's proportional footprint shrinks as competitors grow.

How Does AI Bot Traffic Differ by Country?

The most striking finding in Cloudflare Radar's country-level data is a clean geographic split between two patterns. I'm calling them the "Googlebot Belt" and the "ChatGPT Corridor."

The Googlebot Belt: Western and European Markets

In the United States, United Kingdom, Germany, France, Netherlands, and Canada, Googlebot holds the #1 position among AI crawlers. France stands out as the most Googlebot-heavy country I analyzed, with Googlebot commanding 63.6% of all AI bot traffic -- nearly two-thirds of every AI-related request hitting French websites.

According to Cloudflare Radar regional data, the United States has the highest concentration of pure training crawlers (41.0% of AI bot traffic classified as "Training"), which aligns with the US being home to OpenAI, Anthropic, Meta AI, and Amazon's AI labs. The UK shows an unusually high Bingbot presence at 18.8%, nearly double the global average of 9.7%.

The ChatGPT Corridor: Asia-Pacific and Latin America

In Japan, Brazil, Australia, South Korea, and India, ChatGPT-User -- the bot that visits pages when real users ask ChatGPT questions -- is the dominant AI crawler. This is a fundamentally different pattern. Instead of companies training models, what I'm seeing is end-user demand for AI-powered answers driving the crawl traffic.

Australia is the single most extreme case in my analysis. A staggering 75.8% of all AI bot traffic hitting Australian websites comes from ChatGPT-User, with 76.0% of crawl activity classified as "User Action." This means three out of every four AI crawler requests to Australian sites are triggered by someone actively asking ChatGPT a question -- not by a training pipeline scraping content.

Why Does ChatGPT Dominate in Some Countries but Not Others?

The Googlebot Belt vs. ChatGPT Corridor divide likely reflects three overlapping factors I've identified from the data:

1. AI company headquarters concentration. The US, UK, and Western Europe host the majority of AI companies running training crawlers (OpenAI, Anthropic, Meta, Google DeepMind). Sites in these regions serve content in the languages and domains that training pipelines prioritize.
2. ChatGPT adoption rates. Markets like Brazil, Japan, and Australia have high ChatGPT consumer adoption, but fewer competing AI services. When users in Sao Paulo or Sydney ask ChatGPT a question, the bot fetches content from local websites -- driving the User Action percentage up.
3. Content language and training value. English-language content (US, UK, Canada, Australia) is the most valuable for training. But Australia's small domestic content market combined with high ChatGPT adoption flips the balance toward user-triggered crawling. In contrast, the US has so much English content that training crawlers overwhelm the user action signal.

Which Industries Do AI Bots Target Most?

According to Cloudflare Radar AI Insights (get_ai_data industry dimension), retail and e-commerce sites absorb the largest share of AI crawling activity globally. Shopping-related content is the single most crawled vertical, followed by technology and news media.

The retail dominance at 31.2% makes sense when you consider how AI models are increasingly used for product research and comparison shopping. When a user asks ChatGPT "what's the best running shoe under $150," the bot needs to crawl current product pages and reviews to generate its answer. This is a direct bandwidth cost that e-commerce operators should factor into their infrastructure planning.

What Are AI Bots Actually Doing With the Content They Crawl?

Cloudflare Radar classifies AI bot activity into four purpose categories. Globally, the split reveals that nearly half of all AI crawling serves a dual or ambiguous purpose:

• Mixed Purpose (48.3%): Crawlers like Googlebot that simultaneously index for search and collect training data. This is the largest category because Google and Bing don't cleanly separate their search indexing from their AI data collection.
• Training (42.0%): Dedicated training crawlers like GPTBot, ClaudeBot, and Meta-ExternalAgent that are explicitly collecting data to improve AI models.
• Search (6.9%): AI-powered search bots like OAI-SearchBot that crawl pages to generate search results.
• User Action (2.2%): Bots like ChatGPT-User that fetch pages in real time when users ask questions. Despite being the smallest global category, this is the fastest-growing segment.

The country-level crawl purpose data reveals sharp contrasts. In the United States, 91.6% of AI bot activity is classified as either Mixed Purpose or Training -- meaning US websites serve primarily as raw material for model building. In Brazil, the picture inverts: 57.7% of AI bot activity is User Action, meaning the majority of crawls are triggered by actual people asking AI for answers about Brazilian content.

How Has AI Bot Traffic Changed Month Over Month?

Comparing Cloudflare Radar data from January 9 - February 8, 2026, against the prior 30-day control period (December 10, 2025 - January 9, 2026) reveals several notable shifts in the global AI crawler market:

Meta's training crawler grew its share by 36% in a single month. That's a significant ramp-up and aligns with Meta's public statements about expanding its Llama model training pipeline. OAI-SearchBot's climb from 1.5% to 2.0% is also worth noting -- a 33% relative increase that reflects growing usage of ChatGPT's search feature.

What Should Website Owners Do About AI Bot Traffic?

Based on what I've found in the Cloudflare Radar data, here are specific actions depending on your situation:

The blocking decision itself turns on a single number: how many pages a bot takes for every visitor it sends back. Our crawl-to-refer ratio analysis ranks every major operator on exactly that — a spread of four orders of magnitude, from DuckDuckGo's near-parity to Anthropic's.

If your audience is in the US or Western Europe:

• Expect Googlebot to be your primary AI crawler. Your robots.txt and AI-specific blocking policies will have limited effect on the largest share of AI-related traffic because Google bundles search indexing and AI training under the same bot.
• Monitor Meta-ExternalAgent closely. Its 36% month-over-month growth suggests the crawl volume will continue increasing through 2026.
• GPTBot and ClaudeBot respect robots.txt directives. If you choose to block them, you'll cut approximately 24% of dedicated AI training traffic to your site. Whichever way you decide, validate your robots.txt file afterwards — a malformed directive fails open, and you'll believe you're blocking a crawler that isn't blocked.

If your audience is in Japan, Brazil, Australia, or South Korea:

• ChatGPT-User is your primary concern and opportunity. Blocking it means your content won't appear when users in your market ask ChatGPT questions -- a growing source of referral traffic.
• Training crawlers represent a much smaller share (7-17% of AI traffic vs. 41% in the US), so the "AI is stealing my content" problem is less about training and more about whether you want real-time AI search visibility.
• Optimize for AI citation. Since most of the AI bot traffic in your market is user-triggered, structuring content with clear answers, statistics, and authoritative sources increases the likelihood of being quoted in AI responses. A GEO audit for AI search engines will tell you how well your pages currently hold up on those signals.

For e-commerce and retail sites everywhere:

• Your vertical absorbs 31.2% of all AI bot traffic -- the highest of any industry. Factor AI crawler bandwidth into your infrastructure costs.
• Product pages and comparison content are prime targets. Keep pricing, availability, and specifications current, as AI bots are increasingly pulling this data for real-time user queries.

How I Analyzed This Data

This analysis uses data from Cloudflare Radar's AI Insights endpoint (get_ai_data), which aggregates traffic patterns across Cloudflare's global network spanning 330 cities in 125+ countries. The data covers January 9 through February 8, 2026.

I queried bot traffic breakdowns by user agent, crawl purpose, industry, and vertical across 11 countries: United States, United Kingdom, Germany, Japan, India, Brazil, France, Australia, Canada, South Korea, and Netherlands. Month-over-month comparison uses a 30-day control window (December 10, 2025 - January 9, 2026). All percentages represent share of identified AI bot requests, not share of total web traffic.

Cloudflare's network processes over 81 million HTTP requests per second and 67 million DNS queries per second, providing one of the most comprehensive views of global internet traffic available. The location filter corresponds to the billing country of the Cloudflare customer whose site received the traffic.

Data source: Cloudflare Radar AI Insights - get_ai_data endpoint (radar.cloudflare.com). Last updated: February 8, 2026.