What is GPTBot?
GPTBot is OpenAI's web crawler that collects data from publicly accessible web pages to improve AI models like ChatGPT, while respecting robots.txt and opt-out preferences. Agent Analytics can track when it visits your website.
Overview
| Operated By | OpenAI |
| Source | Official Website |
| Expected To Follow Robots.txt | Yes |
| Insights Last Updated | July 31, 2026 |
Do you operate this agent? Contact us to suggest an update.
Category
Expected Behavior
GPTBot crawls websites to build AI training datasets, and its operator decides which sites, how often, and how deep it goes. Expect broad sweeps that fetch far more pages per visit than a search crawler would, on an unpredictable schedule. Volume can run heavy while a collection pass is underway, then stop entirely.
GPTBot's User Agent
| User Agent | Mozilla/5.0 AppleWebKit/537.36 (KHTML, like Gecko; compatible; GPTBot/1.4; +https://openai.com/gptbot) |
How To Block GPTBot With Robots.txt
Add this rule to your robots.txt file to block GPTBot from accessing your entire website, or use Automatic Robots.txt to block all AI data scrapers at once. You can customize which pages are blocked by swapping out / for a different path.
User-agent: GPTBot # https://knownagents.com/agents/gptbot
Disallow: /Global Insights for GPTBot
As of July 31, 2026, this data reflects agent visits measured across thousands of websites using Agent Analytics, combined with daily scans of the world's top 1000 websites and their robots.txt files.
Robots.txt Blocked Percentage
Country of Origin
Robots.txt Blocking Trend
22% of top websites block GPTBot in their robots.txt files.
Overall AI Data Scraper Traffic
1.7% of all web traffic came from AI data scrapers.
Top Visited Website Categories
The types of websites most frequently visited by GPTBot.
Frequently Asked Questions
Should I Block GPTBot?
Block it if you want control over how your work trains AI models. GPTBot downloads your content into training datasets without attribution, compensation, or any promise of traffic back. The case for allowing it is reach, because models trained on your content can surface your brand in AI answers. Neither choice changes your Google rankings. For comparison, 22% of the top websites we track already have robots.txt rules for GPTBot.
Does GPTBot Follow Robots.txt Rules?
Yes. GPTBot is expected to follow robots.txt rules, so a disallow rule is the right first move. Automatic Robots.txt adds and maintains that rule for you, and Agent Analytics confirms GPTBot actually follows it.
Does GPTBot Access Private Content?
GPTBot targets public content, but the boundary is not always respected. Some training data scrapers go after paywalled or gated pages when the operator wants that data. If a page loads without signing in, assume it can be collected.
Why Is GPTBot Visiting My Website?
Your content matched what OpenAI wants in a training dataset. GPTBot discovers pages mechanically, through links from other sites, sitemaps, and seed lists, not because anyone chose your site personally.
How Can I Tell if GPTBot Is Visiting My Website?
Agent Analytics tracks GPTBot visits and every other known AI agent, crawler, and scraper alongside it, then authenticates each visit against GPTBot's published verification method. You can also check your server logs for requests whose user agent string contains "GPTBot". Look for broad sweeps that fetch large numbers of pages in sequence. A log match is still not proof, because any bot can claim to be GPTBot.