IPs observed
616,648
Flagged abusive
561,095
Abuse ratio
91%
Proxy / VPN
71%
Of the 616,648 IP addresses ffraud has observed in Amazon.com, Inc. (AS16509), 561,095 — 91% — are flagged as abusive. That is an extraordinarily high concentration. Ordinary end-user networks sit in the low single digits. On a dedicated hosting network a rate this high is the classic bulletproof-hosting or proxy-operation signature, where abuse is the business model. On a large shared platform or CDN it instead reflects abusive tenants riding otherwise-legitimate infrastructure. Either way, unfiltered traffic from this network carries real risk. On top of that, 71% of the flagged IPs resolve as proxies or VPN infrastructure — a strong sign the network is being used to launder traffic and hide origin.
AS16509, answered.
Is AS16509 (Amazon.com, Inc.) safe?+
No. AS16509 is high-risk: 91% of the IPs we observe there are abusive. Traffic from this network should be treated with heavy suspicion.
Why does AS16509 have a 91% abuse rate?+
Of the 616,648 IP addresses ffraud has observed in Amazon.com, Inc. (AS16509), 561,095 — 91% — are flagged as abusive. That is an extraordinarily high concentration. Ordinary end-user networks sit in the low single digits. On a dedicated hosting network a rate this high is the classic bulletproof-hosting or proxy-operation signature, where abuse is the business model. On a large shared platform or CDN it instead reflects abusive tenants riding otherwise-legitimate infrastructure. Either way, unfiltered traffic from this network carries real risk. On top of that, 71% of the flagged IPs resolve as proxies or VPN infrastructure — a strong sign the network is being used to launder traffic and hide origin.
How does ffraud measure this?+
The rate is the share of Amazon.com, Inc.'s IPs that ffraud has independently confirmed as abusive, each verified at least twice by our honeypot sensors and community reports. It is a measurement of observed traffic, not a legal judgment of the operator. See how we score at https://ffraud.com/docs/methodology.