NORAD.io
OVERVIEWCOUNTRIESBOTSTHREATS

ANTHROPIC-AI

LOW RISK🤖 AI ASSISTANT BROWSER

Anthropic's AI assistant web browsing capability for real-time information retrieval

ORGANIZATION
Anthropic
FIRST SEEN
2024-03
RESPECTS ROBOTS.TXT
✓ YES
DOCUMENTATION
docs.anthropic.com
—
DAILY VISITS
—
COUNTRIES ACTIVE
TRACKING
STATUS
—
LAST SEEN

📡 ANTHROPIC-AI USER-AGENT STRING

Mozilla/5.0 AppleWebKit/537.36 (KHTML, like Gecko; compatible; Anthropic-AI; +https://anthropic.com)

This is the User-Agent header sent by Anthropic-AI in HTTP requests. Use this to identify Anthropic-AI in your server access logs.

📋 ABOUT ANTHROPIC-AI

Anthropic-AI is the User-Agent identifier used when Claude, Anthropic's AI assistant, accesses the web in real-time during conversations. This is distinct from ClaudeBot, which crawls for training data. Anthropic-AI represents Claude's live web browsing capability, allowing the AI to fetch current information during user interactions.

Like ChatGPT-User for OpenAI, Anthropic-AI makes on-demand page requests driven by user conversations rather than performing systematic crawling. Each request represents a real user interaction where Claude has been asked to look up current information, verify facts, or access web content that extends beyond its training data.

NORAD.io tracks Anthropic-AI separately from ClaudeBot to help site operators distinguish between AI training crawling and real-time assistant browsing. Many sites choose to allow Anthropic-AI (which drives engagement) while blocking ClaudeBot (which collects training data), reflecting different content access policies for different use cases.

🎯 HOW TO DETECT ANTHROPIC-AI

  • ▸User-Agent contains 'Anthropic-AI'
  • ▸Request pattern is sporadic — single page fetches, not systematic crawling
  • ▸Distinct from ClaudeBot which crawls for training data
  • ▸Requests correlate with Claude user activity
  • ▸May fetch specific sections of pages relevant to user queries

🔄 CRAWL BEHAVIOR

On-demand fetching triggered by Claude's web browsing tool use. Single-page requests driven by user conversations. Does not perform systematic crawling. Fetches specific URLs requested during tool use.

PURPOSE

Enables Claude AI to access real-time web information during conversations. Data is used for immediate response generation, not stored for training. Provides Claude with current information beyond its training data cutoff.

🤖 ROBOTS.TXT CONFIGURATION

User-agent: Anthropic-AI
Allow: /

# This is Claude's browsing mode (real-time, not training)
# To block:
# User-agent: Anthropic-AI
# Disallow: /

Anthropic-AI respects robots.txt directives. Add this to your robots.txt file at the root of your domain.

→ Complete Guide: robots.txt for AI Bots

🗺️ WHERE IS ANTHROPIC-AI ACTIVE?

🇺🇸
United States
🇨🇦
Canada
🇲🇽
Mexico
🇬🇹
Guatemala
🇨🇺
Cuba
🇭🇹
Haiti
🇩🇴
Dominican Republic
🇭🇳
Honduras
🇳🇮
Nicaragua
🇸🇻
El Salvador
🇨🇷
Costa Rica
🇵🇦
Panama

⚠️ RELATED THREATS

Prompt Injection

Attempts to override bot instructions via malicious content embedded in web pages

Data Exfiltration

Bots attempting to extract sensitive data from websites including PII and credentials

Credential Stuffing

Automated login attempts using leaked credentials from data breaches

Aggressive Content Scraping

Bots aggressively scraping content beyond robots.txt limits and ToS

🔗 RELATED BOTS

ClaudeBotLOW

Anthropic · Anthropic's web crawler for training Claude AI models

Claude-WebLOW

Anthropic · Anthropic's dedicated web browsing agent for Claude's tool-use web access

📂 MORE 🤖 AI ASSISTANT BROWSERS

ChatGPT-UserOpenAIPerplexityBotPerplexity AIPerplexity-UserPerplexity AIClaude-WebAnthropic

📚 RELATED GUIDES

How to Detect AI Bots →robots.txt for AI Bots →NORAD API Docs →

PROTECT YOUR WEBSITE

Deploy SiteTrust to monitor and control AI bot access to your site with the Agent Passport Standard.

INSTALL SITETRUST →