> For clean Markdown of any page, append .md to the page URL.
> For a complete documentation index, see https://docs.humansecurity.com/llms.txt.
> For AI client integration (Claude Code, Cursor, etc.), connect to the MCP server at https://docs.humansecurity.com/_mcp/server.

# FraudSensor Overview

FraudSensor is HUMAN's post-bid solution for detecting advertising fraud. By analyzing each impression as it occurs, FraudSensor assembles detailed information about the quality of your advertising traffic and identifies different forms of [invalid traffic](/advertising/ivt-taxonomy) (IVT), including [sophisticated IVT](/advertising/ivt-taxonomy#sophisticated-invalid-traffic-sivt), which closely mimics real human behavior.

> **Note**
>
> If you're a [MediaGuard](/advertising/mediaguard-overview) user, FraudSensor's analysis can validate the pre-bid predictions you receive from MediaGuard. To learn more about using MediaGuard and FraudSensor together, see [Closing the Loop](/advertising/closing-the-loop-mediaguard-and-fraudsensor).

## How FraudSensor works

FraudSensor uses [detection tags](/advertising/tag-types-and-environments) to collect a variety of signals about the users driving your ad impressions, including their device characteristics, network information, and behavioral patterns. These signals provide us with enough information to determine whether each impression was valid or invalid. Any invalid impressions are sorted into different [IVT categories](/advertising/ivt-taxonomy) with information about the specific traits displayed by that user.

There are several [types](/advertising/tag-types-and-environments#tag-types) of detection tags that support different ad formats across a variety of environments. However, all detection tags share the same core function: by loading a short code snippet on the device where each ad impression occurs, HUMAN can gather enough signals to construct a detailed picture of that impression. This level of detail helps us detect [sophisticated forms of invalid traffic](/advertising/ivt-taxonomy#sophisticated-invalid-traffic-sivt). You can view your FraudSensor data in the [HUMAN Dashboard](/advertising/understand-the-human-dashboard) or via our [Reporting API](/advertising/reference/about-the-reporting-api). Since FraudSensor performs its analysis after an impression has already occurred, you can't automatically incorporate this post-bid data into real-time bidding workflows; however, you can still use FraudSensor's IVT analysis to assess the quality of your ad traffic, see which traffic sources are driving the highest levels of invalid impressions, and adjust your future bidding criteria accordingly.

[FraudSensor](/advertising/fraudsensor-overview) is designed to work together with [MediaGuard](/advertising/mediaguard-overview). By comparing MediaGuard's pre-bid predictions with FraudSensor's post-bid detection and analysis, HUMAN can create a ["closed-loop](/advertising/closing-the-loop-mediaguard-and-fraudsensor)" ecosystem with data that's richer than the sum of its parts. Therefore the FraudSensor signals are also essential to verify that MediaGuard's predictions are accurate and lets us continuously refine our detection models to identify [sophisticated invalid traffic](/advertising/ivt-taxonomy#sivt).

## Get started with FraudSensor

If you want to get started with FraudSensor, a HUMAN representative will help you build your own unique [detection tag](/advertising/fraudsensor-integration-phases) to serve alongside your ad impressions. These detection tags are tailored to the specific [environment](/advertising/tag-types-and-environments#tag-deployment-environments) where they'll be deployed and can be customized to include a wide variety of tag parameters.

We also recommend setting up an [Impression Sync](/advertising/impression-sync) integration to help identify any bad actors who attempt to evade FraudSensor's detection tag.

To get started, see [FraudSensor Integration Phases](/advertising/fraudsensor-integration-phases).