Skip to main content
Back to glossary

Glossary

anthropic-ai (Bot)

AI Bots

anthropic-ai

anthropic-ai is a secondary user agent from Anthropic (the company behind Claude) used for web crawling. It complements ClaudeBot for training data collection and research purposes.

Why It Matters for GEO

For maximum visibility in Claude, you need to allow TWO bots: ClaudeBot AND anthropic-ai. Missing either one may limit how well your content is indexed.

Many businesses allow ClaudeBot and stop there, not realizing that Anthropic uses this second user agent for additional crawling tasks. Allowing both ensures that no part of your site is missed by Anthropic's data collection.

How to Optimize

Add to your robots.txt:

User-agent: ClaudeBot
Allow: /

User-agent: anthropic-ai
Allow: /

Key Distinction

  • ClaudeBot: Primary bot for Claude products
  • anthropic-ai: Secondary bot for research

Practical Example

A legal services firm wants to appear when corporate clients ask Claude questions about contract disputes. They check their robots.txt and find ClaudeBot is allowed, but anthropic-ai is not listed. They add the missing entry. Since anthropic-ai handles certain types of web research queries, allowing it means their expert content on contract law becomes eligible for a wider range of Claude responses in the legal domain.

Common Mistakes

  • Only allowing ClaudeBot: The most common mistake is treating Anthropic as a single-bot company. Always configure both user agents together.
  • Blocking with a wildcard: Some robots.txt files use User-agent: * with Disallow: / as a default before listing exceptions. If anthropic-ai is not listed as an exception, it is blocked.
  • Forgetting to verify: After editing robots.txt, test it using Google Search Console or a free robots.txt validator. A typo in the user agent name means the rule has no effect.
  • Assuming blocks do not matter for small sites: Anthropic crawls sites of all sizes. Blocking the bot has the same consequence whether you have 10 pages or 10,000.