Dusting for Fingerprints - How it Works
Fake reviews can be impossible to spot by eye, but they do leave a detectable fingerprint.
A.I. Generated Fake Reviews
Today, Generative AI has made it possible to fake reviews at unprecedented scale and review faking services are widely available at low cost.
However, suspicious reviews leave a fingerprint, and these emit signals.
The more obvious fingerprints include spikes in highly positive or negative reviews in a short space of time, generic reviews without specific detail about a product, reviews that use very similar language as other reviews for the same product, numerous reviews written from the same username or email address.
Even these fingerprints are often invisible to the human eye, particularly when the suspicious reviews are buried deep among genuine reviews, or they make up the vast majority of pages and pages of reviews.
The TruthEngine detects all of these characteristics plus many, many more.
We Spot the Signals
The TruthEngine is an AI-powered system that works at scale, rapidly and accurately analyses large numbers of reviews stretching back in time. Fake reviews are often impossible to detect by eye, particularly for consumers having a quick look at a few reviews before they decide to buy something.
But, Fake Reviews do Leave a Detectable Fingerprint.
The TruthEngine uses machine learning and proprietary pattern recognition algorithms to rapidly and accurately detect changes in the characteristics and makeup of a company’s reviews. We’ve trained it to detect and flag the signals of fake and manipulated reviews - whether they are positive or negative.
It is language agnostic, multi-platform and privacy compliant. It works automatically, remotely and at scale.
And it’s built to run repeatably to help businesses stay compliant.