Amazonbot
The Amazon crawler feeding Alexa answers and, per Amazon, the training of its AI services. A steady mid-volume presence in most server logs. Registry class is crawler because its index serves live Alexa responses, though it doubles as a training source.
What it feeds
Amazon describes Amazonbot as crawling to improve its products and services,[1] and states the data may be used to train Amazon AI models.[2] In practice that covers Alexa answers as well as model training, which is why the registry files it as a crawler rather than a pure training scraper.
How to identify it
The user agent is a browser-shaped string naming the crawler and its version.[3] Amazon publishes the addresses it crawls from at developer.amazon.com/amazonbot/ip-addresses, and that list, not the name, is what separates a real fetch from an impersonation.
Blocking and its consequences
Amazon states the crawler respects the robots exclusion protocol and its allow and disallow directives.[4] A robots.txt change is not immediate: Amazon documents fetching host-level robots.txt or using a copy cached within the last 30 days,[5] so expect up to a month of lag before a new rule bites.
Verification
Operator documentation: https://developer.amazon.com/support/amazonbot
A user agent is a claim, not a proof: any client can send any name. Where an operator publishes IP ranges, a reverse lookup is the only way to tell a real visit from an impersonation.
Seen by Baseline
Requests from this bot across sites tracked by Baseline, last 30 days. Aggregated over - sites, never reported per site.
Last updated 2026-09-05. Written and maintained by Baseline Labs.
