Does science.org block AI crawlers?
As of 2026-11-03, the robots.txt of science.org blocks 5 of 18 AI crawlers, including GPTBot, ChatGPT-User, Google-Extended, Applebot-Extended.
AI crawler status of science.org
Based on the live robots.txt of science.org, checked on 2026-11-03. “Some paths restricted” means the bot may crawl the site but certain sections are excluded.
| User agent | Company | Purpose | Status |
|---|---|---|---|
GPTBot | OpenAI | Model training | Blocked |
OAI-SearchBot | OpenAI | ChatGPT search index | Allowed (some paths restricted) |
ChatGPT-User | OpenAI | Live browsing for ChatGPT users | Blocked |
ClaudeBot | Anthropic | Model training | Allowed (some paths restricted) |
Claude-SearchBot | Anthropic | Claude search index | Allowed (some paths restricted) |
Claude-User | Anthropic | Live browsing for Claude users | Allowed (some paths restricted) |
PerplexityBot | Perplexity | Perplexity search index | Allowed (some paths restricted) |
Perplexity-User | Perplexity | Live browsing for Perplexity users | Allowed (some paths restricted) |
Google-Extended | Gemini training & grounding | Blocked | |
Googlebot | Search and AI Overviews | Allowed (some paths restricted) | |
Bingbot | Microsoft | Bing search and Microsoft Copilot | Allowed (some paths restricted) |
Applebot-Extended | Apple | Apple Intelligence training | Blocked |
Meta-ExternalAgent | Meta | Meta AI training | Allowed (some paths restricted) |
Amazonbot | Amazon | Alexa and Amazon AI | Allowed (some paths restricted) |
DuckAssistBot | DuckDuckGo | DuckAssist answers | Allowed (some paths restricted) |
MistralAI-User | Mistral | Live browsing for Le Chat users | Allowed (some paths restricted) |
CCBot | Common Crawl | Open dataset used by many AI models | Blocked |
Bytespider | ByteDance | Model training | Allowed (some paths restricted) |
What this means
science.org follows a pattern many publishers now use: it keeps AI training crawlers like GPTBot, Google-Extended, CCBot out, but still lets AI search and live-browsing agents in. That way it can still be cited and linked in ChatGPT or Perplexity answers.
Among the 92 reference & knowledge sites in our AI crawler report, 27% block GPTBot completely. science.org is one of them.
Blocking AI bots is one side. The other: is AI actually sending you visitors? Citelytic shows how much traffic ChatGPT, Perplexity, Gemini and Copilot send to your own site.
Start free – no credit cardCheck your own website
Want to see this table for your own domain? The free AI crawler checker reads your robots.txt and shows which AI assistants can access your site. For a full 0-100 rating with fixes, get your AI visibility score. No signup needed.
FAQ
Does science.org block ChatGPT?
Partly. As of 2026-11-03, OAI-SearchBot is allowed with some restricted paths and ChatGPT-User is blocked in the robots.txt of science.org. GPTBot (OpenAI's training crawler) is blocked.
Can Perplexity access science.org?
PerplexityBot is allowed with some restricted paths and Perplexity-User is allowed with some restricted paths on science.org according to its robots.txt checked on 2026-11-03.
Does science.org have an llms.txt file?
No, we did not find an llms.txt file on science.org when we checked on 2026-11-03.
How is this data collected?
We fetch the public robots.txt file of each site and evaluate the rules for 18 AI user agents for the site root. robots.txt is a request, not a technical barrier, and sites may also block bots at the firewall level, which this check does not see.
Blocking AI bots is one side. The other: is AI actually sending you visitors? Citelytic shows how much traffic ChatGPT, Perplexity, Gemini and Copilot send to your own site.
Start free – no credit card