How to Block DuckAssistBot With Robots.txt (2026)

DuckAssistBot is DuckDuckGo's crawler for AI-assisted answers. It fetches pages in real time when Duck.ai needs to answer a question, and it cites the sources it uses. If you do not want your content appearing in those answers, you can opt out with three lines in robots.txt — and, importantly, opting out will not touch your DuckDuckGo search rankings. Here is the exact rule, what it does, and when you should (and should not) use it.

User-agent: DuckAssistBot
Disallow: /

That is the whole opt-out. Once you add it, DuckDuckGo says the change takes effect after 72 hours, and DuckAssistBot will stop crawling your site. The rest of this guide explains what you are actually opting out of, how DuckAssistBot differs from DuckDuckBot, and how to check that the block is working.

What DuckAssistBot Actually Does

DuckDuckGo runs two crawlers that site owners constantly confuse. Keeping them straight is the key to this entire decision:

• DuckDuckBot is DuckDuckGo's search crawler. It indexes pages for ordinary DuckDuckGo search results. Blocking it removes you from DuckDuckGo search.
• DuckAssistBot is the crawler for AI-assisted answers. It fetches pages in real time when Duck.ai is building an answer, so the answer can cite current sources.

DuckAssistBot's job is retrieval, not training. DuckDuckGo states that the data DuckAssistBot collects is not used to train AI models. Its purpose is to ground an answer in real, citable pages — which is why answers built with it show sources. When you block DuckAssistBot, you are removing your site as a potential source for those AI-assisted answers. You are not opting out of model training, because DuckAssistBot was never training a model in the first place.

That distinction matters. Many site owners block bots in a panic about training data and accidentally give up citation traffic they would have wanted. DuckAssistBot is one of the "good citizen" AI crawlers: it identifies itself clearly, it honors robots.txt, and it sends cited traffic back. Before you block it, make sure you know which of those two things you are giving up.

How to Block DuckAssistBot With Robots.txt

Add this block to your robots.txt file, then upload it to your site root:

User-agent: DuckAssistBot
Disallow: /

A few rules to get right:

• Use the exact token. The user agent is DuckAssistBot. DuckDuckGo publishes the full string as DuckAssistBot/1.2; (+http://duckduckgo.com/duckassistbot.html), but in robots.txt you match on the token DuckAssistBot.
• Keep it in its own group. Each User-agent line starts a group, and the Disallow lines beneath it apply to that group. Do not stack DuckAssistBot into a long list of user agents unless you intend every token in that list to get the same rules.
• Do not touch DuckDuckBot. Unless you also want out of DuckDuckGo search results, leave DuckDuckBot alone. The two tokens are independent: blocking one has no effect on the other.
• Keep your sitemap line. Your robots.txt should still end with your Sitemap: line so search crawlers can discover your URLs.

If you would rather generate the file than hand-write it, use the free robots.txt generator and add DuckAssistBot to your block list, then check the result with the robots.txt tester before you publish.

What Happens After You Block It

According to DuckDuckGo's own help documentation, once you update robots.txt to disallow DuckAssistBot, the change takes effect after 72 hours, and DuckAssistBot will stop crawling your site. Three consequences follow:

1. Your pages stop being sources for AI-assisted answers. Duck.ai will not fetch your pages when building answers, and your site will not be cited there.
2. Your organic rankings do not change. DuckDuckGo is explicit: opting out of DuckAssistBot does not impact organic search rankings and will not affect whether your site appears in DuckDuckGo search results. The search index (DuckDuckBot) and the answer fetcher (DuckAssistBot) are separate systems.
3. Existing answers may linger briefly. Answers generated before the opt-out took effect were built from pages fetched earlier. The opt-out stops future fetching; it does not rewrite history.

If the block does not seem to work after the 72-hour window, DuckDuckGo invites publishers to contact them at crawling@duckduckgo.com with opt-out problems. Before you write, double-check that your robots.txt is reachable at yourdomain.com/robots.txt, that the token is spelled correctly, and that no broader Allow rule is overriding your block.

Should You Block DuckAssistBot or Allow It?

This is a business decision, not a technical one. Here is how to think about it:

Reasons to allow DuckAssistBot

• Cited traffic. Duck.ai answers prominently cite their sources. Being a cited source puts your site in front of people who are actively asking questions in your niche — visitors who arrive pre-qualified and curious.
• No training risk. DuckDuckGo says this data is not used to train models. If training is your main concern, DuckAssistBot is not the bot you need to worry about.
• It behaves well. It identifies itself, honors robots.txt, and publishes its user-agent string and opt-out process. These are the traits of a crawler you can live with.

Reasons to block DuckAssistBot

• You do not want your content in AI answers at all. Some publishers — paid newsletters, licensed databases, sites with content they sell — do not want any AI product summarizing or quoting their pages, even with a citation.
• You are applying a blanket AI policy. If your policy is "no AI products use our content," then DuckAssistBot goes on the block list along with the training crawlers, even though its purpose is different.
• Server load. Real-time answer fetchers can generate bursts of requests when a topic trends. On a small site on shared hosting, that load is a legitimate reason to opt out.

Our general advice: if you publish free content and want traffic, allow DuckAssistBot. If you sell the content itself, block it. The worst choice is blocking it by accident while trying to block something else — which is exactly what happens when people copy a giant block list without reading it.

How to Verify the Block Is Working

1. Check the file. Open yourdomain.com/robots.txt and confirm the DuckAssistBot group is present and correct.
2. Test the rule. Paste your robots.txt into the robots.txt tester and test the DuckAssistBot user agent against a few of your URLs.
3. Wait 72 hours. DuckDuckGo's stated window. Do not declare failure on day one.
4. Check your server logs. After the window passes, search your access logs for DuckAssistBot. If you still see visits, confirm the requests are not coming from cached behavior, then check for a conflicting rule — a wildcard Allow group above your block, a typo in the token, or a second robots.txt served by a CDN or plugin.

Note that server logs are the only reliable proof. DuckDuckGo's help pages do not offer a publisher dashboard for DuckAssistBot, so your logs are your source of truth.

Common Mistakes

• Blocking DuckDuckBot by accident. The two names differ by a few letters, and blocking the wrong one removes you from DuckDuckGo search. Double-check the token before you save.
• Copying a block list that includes DuckAssistBot without realizing it. Many "block all AI bots" lists include it. If you want Duck.ai citations, remove that one line.
• Assuming robots.txt removes you from AI answers instantly. The 72-hour window is real. Be patient, then verify in logs.
• Expecting robots.txt to stop every AI product. Each provider has its own tokens. Blocking DuckAssistBot does nothing about GPTBot, ClaudeBot, PerplexityBot, or any other provider's crawlers. For the full picture, see the complete list of AI crawlers and our guide on getting cited by ChatGPT — the allow side of the same coin.

DuckAssistBot vs Other AI Answer Crawlers

DuckAssistBot is not the only crawler that fetches pages for AI answers, and the differences between them matter for your blocking policy. PerplexityBot builds the index behind Perplexity search, while Perplexity-User fetches pages live when a user asks a question. ChatGPT has a similar split: OAI-SearchBot powers ChatGPT search, and ChatGPT-User fetches pages on demand. Each provider also runs training crawlers, such as GPTBot and ClaudeBot, which are separate tokens with a separate purpose.

The key point is that blocking one provider does nothing about the others. A site that blocks DuckAssistBot but leaves GPTBot and PerplexityBot open is still feeding OpenAI training and Perplexity search. That may be exactly what you want, or it may be an accident. Either way, build your policy provider by provider, not bot by bot, and write it down.

It also helps to separate two decisions that people often mix up. The first is training: do you want a provider to use your content to train models. The second is answers: do you want a provider to fetch and cite your pages in live answers. DuckAssistBot sits squarely in the answers category, with no training involved. GPTBot sits in the training category. Many publishers block training crawlers while allowing answer crawlers, because answers bring cited traffic and training does not. If that is your policy, DuckAssistBot belongs on your allow list, not your block list.

FAQ

What is DuckAssistBot?

DuckAssistBot is DuckDuckGo's web crawler for AI-assisted answers on Duck.ai. It crawls pages in real time so answers can cite current sources. Its user agent appears as DuckAssistBot/1.2; (+http://duckduckgo.com/duckassistbot.html).

Does DuckAssistBot train AI models on my content?

No. DuckDuckGo states that the data DuckAssistBot collects is not used in any way to train AI models. It is used to build cited answers at the moment a question is asked.

Will blocking DuckAssistBot hurt my DuckDuckGo rankings?

No. DuckDuckGo says opting out of DuckAssistBot does not impact organic search rankings and does not affect whether your site appears in DuckDuckGo search results. Search indexing is handled by the separate DuckDuckBot crawler.

How long does the DuckAssistBot opt-out take?

72 hours after you update robots.txt. After that window, DuckAssistBot should stop crawling your site. Verify in your server logs.

Can I block DuckAssistBot on WordPress, Shopify, or Wix?

Yes — anywhere you can control robots.txt. On WordPress, edit the file or use an SEO plugin. On Shopify, edit the robots.txt template. On Wix, use the robots.txt editor. The rule itself is identical everywhere: User-agent: DuckAssistBot plus Disallow: /.

What if DuckAssistBot keeps visiting after 72 hours?

First re-check your robots.txt for typos and conflicting rules. If the file is correct, contact DuckDuckGo at crawling@duckduckgo.com — they publish that address specifically for opt-out problems.

Bottom line: DuckAssistBot is one of the better-behaved AI crawlers: it cites sources, does not train models on your content, and honors a simple robots.txt opt-out within 72 hours — without touching your search rankings. Block it if you do not want your content in AI answers at all. Allow it if you want cited traffic from Duck.ai. Either way, make the choice on purpose, and do not confuse it with DuckDuckBot.