Grammarly AI: Web Sources, Crawlers and Site Access
Grammarly is an AI writing assistant, and its company is now called Superhuman. Grammarly compares text with web pages, suggests web sources for citations and answers questions about the page that a user reads. Grammarly documents no crawler, user agent or robots.txt token, so a site owner has no specific rule to add.
Grammarly and Superhuman
The Grammarly about page says that Grammarly is now part of the Superhuman Suite. The company acquired Coda and Superhuman Mail in 2025. The page then says: “In October, our company became Superhuman.” The Grammarly help center also covers Superhuman Go, an AI assistant that works in the browser extension.
How Grammarly uses web pages
Three Grammarly features use content from the web:
- Plagiarism Checker. It compares text with web pages and academic papers. The plagiarism checker page says that it checks against over 16 billion web pages and the academic databases of ProQuest. The user guide says that it marks matching parts of the text with bars.
- Citation Finder. It flags claims that need a citation and suggests links to web-based sources (Citation Finder user guide).
- Reading assistance in Superhuman Go. A user selects text on a web page and asks a question. The answer is based on what is on the screen of the user (reading assistance).
Grammarly does not say how it collects the web pages for the plagiarism checker or for citation suggestions.
What a site owner can control
A user agent is the client name in the User-Agent HTTP header. As of October 2026, the Grammarly website and help center document no crawler, user agent, IP list or robots.txt token. So no documented robots.txt rule removes your pages from plagiarism checks or citation suggestions.
Reading assistance works on a page that a person already opened in the browser. robots.txt gives rules to crawlers (RFC 9309). It does not control what a browser extension reads from an open page.