How detection works
Every visit is tied to one browser on one device through a persistent visitor identity that survives IP changes. Device, behavior and network signals score that visitor; a different browser or device is a different person, and clearing site data resets the identity.
On this page
Signals, scores and identity
- Device signals
- Canvas, audio, GPU and TLS characteristics are collected as fraud evidence: hardware that mutates under one visitor id, or a fresh browser every session from one IP, is a signal. They never define who a person is, and a fingerprint is not a verified real-world identity.
- Behavior signals
- Engagement and repeated activity help assess the visit. Short visits can also come from legitimate customers.
- Network signals
- IP and network history add context, but shared networks require care.
- Visitor id
- One browser on one device. The id is issued by the AdProtektor server on the first visit and survives IP changes; clearing site data resets it, and a different browser or device is a different visitor.
Automatic action
Configure protection for each website, then let AdProtektor evaluate traffic and apply eligible actions automatically. Open the dashboard when you want to inspect reasons, evidence or exceptions; no per-visitor approval is required.
AI analysis and its influence on paid-click blocking, website blocking and threat scoring are separate settings. Eligible pipeline verdicts can affect enforcement only when the relevant controls and safety gates allow it. Opening an AI explanation in the dashboard does not itself authorize a block.
What recognition cannot promise
Clearing cookies or changing networks does not necessarily erase all available signals. It also does not guarantee continued recognition. A fresh browser or device can need a new evaluation.
False positives are possible. Inspect the reason and available evidence, adjust the relevant settings and whitelist a known legitimate visitor when appropriate.
Frequently asked questions
Can visits be linked when the IP changes?
Yes. The visitor identity belongs to the browser, not the address, so the same browser keeps the same id across IP, VPN and network changes. An IP change alone neither proves nor disproves fraud.
Does clearing cookies make someone unrecognizable?
Clearing site data resets the visitor identity, so the next visit starts as a new visitor. Fingerprint and network signals are still collected as evidence — a fresh browser every session from one IP is itself a pattern the scoring looks for — but they do not re-link the visits.