Bots
What is Amzn-User, and should you allow it?
Amzn-User is Amazon's user-triggered fetcher: when someone asks Alexa a question that needs current information, it may fetch your page live to answer. Allow it; it means a person's question reached you, and Amazon says it does not crawl for model training. Amazon also says it may not follow all robots.txt directives, so a rule is only a preference.
7 days free · card required
0requests filed as Amzn-User · 30 days
of 143,601 AI crawler requests from 7 operators
| Agent | AI requests, 30 days | attensira.com robots.txt |
|---|---|---|
| Amzn-User (this page) | 0 | Wildcard |
| ChatGPT-User | ≈230 | Allowed |
| Claude-User | ≈36 | Allowed |
| MistralAI-User | 0 | Allowed |
| Perplexity-User | 0 | Allowed |
2 of 5 had requests filed in the window. User agents as sent, not checked against address lists.
Analyst · reads the AI traffic boardAI traffic board, read over the Attensira MCP, and our live robots.txt · · n=143,601 AI requests
Karl-Gustav Kallasmaa, Founder & CEOLast updated How do I allow or block Amzn-User in robots.txt?
Paste this into the robots.txt at the root of your domain. The operator says the agent may not read robots.txt, so the group records your preference; enforce anything sensitive on the server. Beside it, the group attensira.com serves Amzn-User today.
# Addressable, but Amazon documents this fetcher as# possibly not following all robots.txt directives. Treat# it as a preference.User-agent: Amzn-UserDisallow: /internal/# No group names Amzn-User, so it reads the wildcard# group.User-agent: *Allow: /Disallow: /api/We do not name Amzn-User. It falls through to the wildcard group, which allows everything except /api/. Its operator states robots.txt may not apply to it, so this group is a preference at most.
Read the whole file, or write yours with the robots.txt generator.
Who runs Amzn-User, and what is it for?
Amazon runs it, and we file it under user-triggered fetch. Fetches one page live when a person asks Alexa a question that needs current information. Amazon says it does not crawl for model training, and that it may not follow every robots.txt directive.
From Amazon’s documentation, read . Why a training fetch and a search fetch are different decisions: the AI crawler guide.
Why should you let Amzn-User in?
Because each request is a person's question reaching your page. Amazon documents Amzn-User as supporting user actions, such as responding to Alexa queries that require up-to-date information: when a customer asks, it may fetch live information from the web to give an accurate answer on the user's behalf1.
The Alexa framing is what makes this token unusual. Most user-triggered fetchers sit behind a chat box, where a link in the answer is part of the product. A voice assistant reads the answer out. Your page is still consulted, but visible attribution may be weaker or absent. That is worth knowing; it is not a reason to refuse the reader.
Amazon also states that Amzn-User does not crawl content for generative AI model training3. Of Amazon's three independent tokens, only Amazonbot carries the wording that content may be used to train Amazon AI models. A site owner objecting to training has one token to name, not three. Naming all three costs live answers and Alexa eligibility for no additional protection.
Does a robots.txt rule stop Amzn-User?
Not reliably, and Amazon says so: because actions taken by Amzn-User can be initiated by a user, it may not follow all robots.txt directives2. This page records its position as not bound.
The token is still addressable. Amazon lists it among the three settings, so a rule expresses a preference and costs nothing. The robots.txt block above scopes it to an internal path and says in a comment that it is a preference. Treating it as a wall produces a configuration that looks safe and is not. If a path must not be read by an assistant, put it behind authentication.
There is one enforceable page-level lever. Amazon states its agents respect rel=nofollow at link level and the noarchive, noindex and none robots meta tags, and that noarchive means do not use the page for model training6. That governs use, not fetching.
How do the user-triggered fetchers compare?
| Fetcher | Operator's robots.txt position |
|---|---|
| Amzn-User | May not follow all robots.txt directives |
| ChatGPT-User | Robots.txt rules may not apply |
| Perplexity-User | Generally ignores robots.txt |
| Claude-User | Honours robots.txt, per Anthropic |
Most operators have converged on one line: robots.txt governs crawling, and a fetch a human just asked for is not crawling. Anthropic is the exception. Whether the line convinces you is a fair argument; what it is not is undisclosed. The Perplexity pair is set out in PerplexityBot vs Perplexity-User.
Why does verification matter more for Amzn-User?
Because sites exempt user-triggered traffic from rate limits, which makes the name worth impersonating. Amazon publishes this agent's addresses on its live address page, separate from the two crawler lists5. Amazon calls it the live list, which fits. If a burst of Amzn-User requests after a launch fails that check, it is not Amazon, and a network block is both justified and the only thing that works.
What do verified Amzn-User fetches tell you?
That Amazon's assistant consulted your pages to answer a live customer question. Like every user-triggered token, it produces no analytics session and may produce no click. It is still the earliest observable trace of being in an answer path, so it belongs in the citation-capable column of your log analysis, not in bot noise.
- IP range file
- live-ip-addresses
Amazon publishes no reverse-DNS suffix for Amzn-User; verify by the range file.
ResearcherTranscribed from Amazon documentation · How should you match it in logs?
Amazon's example header is Amzn-User/0.14, with W.X.Y.Z as a literal placeholder. Match the bare token Amzn-User, case-insensitively, as a whole word. A filter for Amzn- merges it with Amzn-SearchBot and loses the difference between a customer waiting for an answer and routine indexing.
What can Attensira show you about Amzn-User?
AI traffic, on Growth and Business, counts which pages AI agents fetch, grouped by operator and declared purpose, measured at the edge, so live-answer fetches are counted apart from crawls. It reports user agents as sent; the live-list check stays yours. See crawler logs and the robots.txt generator.
Mozilla/5.0 AppleWebKit/537.36 (KHTML, like Gecko; compatible; Amzn-User/0.1) Chrome/W.X.Y.Z Safari/537.36Agents filed beside Amzn-User
Should I allow Amzn-User in my situation?
Concrete cases, and what each one implies for this agent.
You disallowed Amzn-User and the fetches continued.
Accept that robots.txt is a preference here, and decide whether the network-layer block is worth its cost.
Amazon documents that Amzn-User may not follow all robots.txt directives because a user initiated the action. Enforcing the block means a customer asking Alexa about you gets an answer assembled without your page — which for most sites is a worse outcome than the fetch.
You are trying to work out whether Amazon traffic is training on you or answering questions about you.
Split the logs three ways by token before drawing any conclusion.
Amazon documents three independent tokens with three different jobs, and states that both Amzn-User and Amzn-SearchBot do not crawl content for generative AI model training. Only Amazonbot carries the may-be-used-for-training wording. A single Amazon row in a dashboard cannot answer the question.
A burst of Amzn-User requests arrives after a product launch and you suspect scraping.
Verify against the live address list before rate-limiting.
A user-triggered token is attached to real people asking real questions, which is exactly why it is worth borrowing. If the addresses are not on Amazon's list, the header is a costume, and throttling the genuine fetcher would have degraded live answers about your launch for no protective benefit.
What else do people ask?
Does Amzn-User obey robots.txt?
Is Amzn-User training Amazon's AI models?
Can I stop Amazon using a page for training if Amzn-User fetches it?
Where do these facts come from?
6 claims on this page, each with the page it was read from; last checked . The numbers in the text point here.
Amazon states that Amzn-User supports user actions, such as responding to Alexa queries that require up-to-date information, and may fetch live information from the web to provide accurate answers on the user's behalf.
developer.amazon.com · retrieved
“Amzn-User supports user actions, such as responding to Alexa queries that require up-to-date information.”
Amazon states that because actions taken by Amzn-User can be initiated by a user, it may not follow all robots.txt directives.
developer.amazon.com · retrieved
“Because actions taken by Amzn-User can be initiated by a user, it may not follow all robots.txt directives.”
Amazon states that Amzn-User does not crawl content for generative AI model training, and documents Amazonbot, Amzn-SearchBot and Amzn-User as three independent user-agent settings.
developer.amazon.com · retrieved
“Amzn-User does not crawl content for generative AI model training.”
Amazon gives Amzn-User's example user-agent string as "Mozilla/5.0 AppleWebKit/537.36 (KHTML, like Gecko; compatible; Amzn-User/0.1) Chrome/W.X.Y.Z Safari/537.36".
developer.amazon.com · retrieved · changes often, check the source
Amazon publishes Amzn-User's addresses at developer.amazon.com/amazonbot/live-ip-addresses/, separately from the Amazonbot and Amzn-SearchBot lists.
developer.amazon.com · retrieved
Amazon states that when these user agents access web pages they respect the link-level rel=nofollow directive and the page-level robots meta tags noarchive, noindex and none, with noarchive meaning do not use the page for model training.
developer.amazon.com · retrieved
Want whatever it takes to win the answer?
7 days free · card required