How Amazon is using AI to detect fake product reviews and ensure authentic customer feedback
Discover how Amazon utilizes sophisticated AI to publish genuine feedback and filter out fraudulent content.
Customer reviews have remained a fundamental aspect of the Amazon shopping experience since the company’s inception in 1995. We ensure that it is simple for shoppers to provide honest feedback, which helps millions of people worldwide make informed purchasing decisions. Simultaneously, we work to prevent bad actors from exploiting our trusted marketplace. This is where artificial intelligence (AI) plays a vital role.
So, what occurs when a customer submits a review? Before any content goes live, we employ AI to scan for known indicators of fraud. The vast majority of submissions meet our high standards for authenticity and are published immediately. However, if we detect potential abuse, we take several steps. If we are certain a review is fake, we act quickly to block or remove it and pursue further measures, such as revoking review privileges, banning malicious accounts, or even initiating litigation against the involved parties. If a review appears suspicious but requires further verification, our expert investigators—who are specifically trained to spot abusive behavior—examine additional signals before taking action. In 2023, we proactively blocked more than 250 million suspected fake reviews from our stores worldwide.
“Fake reviews intentionally mislead customers by providing information that is not impartial, authentic, or intended for that <a href="https://shoppixy.com/online-product-reviews-are-becoming-a-battlefield-for-modern-ai/” title=”Online product reviews are becoming a battlefield for modern AI”>product or service,” says Josh Meek, Senior Data Science Manager on Amazon’s Fraud Abuse and Prevention team. “Not only do millions of customers count on the authenticity of reviews on Amazon for purchase decisions, but millions of brands and businesses count on us to accurately identify fake reviews and stop them from ever reaching their customers. We work hard to responsibly monitor and enforce our policies to ensure reviews reflect the views of real customers, and protect honest sellers who rely on us to get it right.”
Beyond these measures, we leverage the latest AI advancements to intercept hundreds of millions of suspected fraudulent reviews, manipulated ratings, fake accounts, and other forms of abuse before they ever reach the customer. Our machine learning (ML) models evaluate a wide range of proprietary data, including whether a seller has invested in advertising that might be driving a surge in reviews, reports of abuse submitted by customers, risky behavioral patterns, and historical review data.
Large language models (LLMs) are used in conjunction with natural language processing to detect anomalies that might suggest a review is fraudulent or incentivized by gift cards, free products, or other forms of compensation. We also utilize deep graph neural networks (GNNs) to analyze complex relationships and behavioral patterns, which helps us identify and remove networks of bad actors or flag suspicious activity for further investigation.
Our Senior Data Science Manager, Josh Meek says: “The difference between an authentic and fake review is not always clear for someone outside of Amazon to spot. For example, a product might accumulate reviews quickly because a seller invested in advertising or is offering a great product at the right price. Or, a customer may think a review is fake because it includes poor grammar.”
This is where some of our critics get fake review detection wrong—they have to make big assumptions without having access to data signals that indicate patterns of abuse. By combining advanced technology with proprietary data, we can identify fraudulent reviews more accurately, looking past surface-level indicators to uncover deeper connections between bad actors.
“Maintaining a trustworthy shopping experience is our top priority,” says Rebecca Mond, Head of External Relations, Trustworthy Reviews at Amazon. “We continue to invent new ways to improve and stop fake reviews from entering our store and protect our customers so they can shop with confidence.”
