Does pnas.org block AI crawlers?
Blocks AI training, allows AI answers
pnas.org blocks at least one AI training crawler but still lets AI answer engines read its pages.
Checked from pnas.org/robots.txt. Robots rules change; the live file is the source of truth.
- AI crawlers blocked
- 1 of 12
- AI bots named in robots.txt
- 1
- llms.txt
- Not found
- Popularity
- Tranco top 2,000 (#1,983)
Can AI assistants cite pnas.org?
- ChatGPT search: its search crawlers may read the site
- Claude: its search crawlers may read the site
- Perplexity: its search crawlers may read the site
pnas.org opts out of AI training by OpenAI. Training crawlers and answer crawlers are separate: blocking training doesn't by itself remove a site from AI answers.
AI crawler access, bot by bot
| Crawler | Used for | Status on pnas.org |
|---|---|---|
| GPTBotOpenAI | AI training | Blocked |
| OAI-SearchBotOpenAI | AI answers | Some paths restricted |
| ChatGPT-UserOpenAI | AI answers | Some paths restricted |
| ClaudeBotAnthropic | AI training | Some paths restricted |
| Claude-SearchBotAnthropic | AI answers | Some paths restricted |
| Claude-UserAnthropic | AI answers | Some paths restricted |
| PerplexityBotPerplexity | AI answers | Some paths restricted |
| Google-ExtendedGoogle | AI training | Some paths restricted |
| Applebot-ExtendedApple | AI training | Some paths restricted |
| CCBotCommon Crawl | AI training | Some paths restricted |
| BytespiderByteDance | AI training | Some paths restricted |
| Meta-ExternalAgentMeta | AI training | Some paths restricted |
How pnas.org compares
- 19% of the 1,278 popular sites we checked block GPTBot; pnas.org is one of them.
- 18% of the 1,278 popular sites we checked block ClaudeBot; pnas.org does not.
- 14% of the 1,278 popular sites we checked block PerplexityBot; pnas.org does not.
- Among 162 sites of similar popularity, 25% block at least one AI crawler.
The robots.txt rules that apply to AI crawlers
User-agent: * Disallow: /action Disallow: /search Disallow: /media Disallow: /servlet/linkout Disallow: /author/ Disallow: /authored-by/ Disallow: /personalize/ Disallow: /doi/metrics Disallow: /doi/citedby Disallow: /doi/reader Disallow: /doi/ref Disallow: /doi/epdf/ Disallow: /history/ Allow: /action/showJournal Allow: /action/showXml Allow: /action/showFeed Allow: /action/showfeed User-agent: GPTBot Disallow: /
Excerpt: only groups for all user-agents (*) or named AI crawlers, up to 40 rules per group.
Sites with a similar AI policy
- diretta.it · 3 blocked
- euronews.com · 3 blocked
- strava.com · 4 blocked
- oreilly.com · 4 blocked
- meetup.com · 1 blocked
- gizmodo.com · 3 blocked
- merkur.de · 7 blocked
- 1c.ru · 1 blocked
All sites that are “blocks ai training, allows ai answers” →
Does pnas.org block GPTBot?
Yes. As of Oct 2, 2026, pnas.org's robots.txt disallows GPTBot, OpenAI's training crawler.
Can ChatGPT search show results from pnas.org?
Its robots.txt doesn't block OAI-SearchBot or ChatGPT-User, so ChatGPT's search crawlers may read it. Whether pages are actually shown depends on ChatGPT.
Does pnas.org have an llms.txt file?
We didn't find a valid llms.txt at pnas.org/llms.txt when we checked.
Data: pnas.org's public robots.txt and llms.txt, fetched by AskableHQ. Popularity rank from the Tranco list (ID Y83YG). Methodology.