Amazon Cracks Down on Fake Reviews Using AI

According to Amazon, the ‘fake review spotting’ AI tool uses machine learning to find relations between a purchasing account and the seller

If you are an avid online shopper, reviews are what you trust, they give you a better detail about the quality and type of product that you are going to receive before you have it in your own hand, but what if we tell you that at least 4% to 30% of the reviews posted on every e-commerce store or marketplace are fake.

Sounds shocking right? It sure does; it also means that a lot of our shopping decisions we make come after reading fake reviews posted by the seller itself.

Ecommerce platforms, both big and small, have been trying to counter this problem for years but sellers have always found newer and foolproof ways to continue benefiting from fake reviews and comments.

Fake reviews can be easily spotted if someone has a very high percentage of five star reviews, but it’s almost impossible to judge them otherwise on a larger scale.

Amazon, the world’s biggest e-commerce company, however has now harnessed the powers of AI to tackle fake reviews and will use this new technology to identify reviews and comments that are not genuine.

The company has spent a lot of its fortune towards machine learning models which analyze a number of different data points to find any fraudulent behavior.

Amazon claimed that its fraud-detection AI could assess several variables to determine the possibility that a review is fake.

This may include the author's connections to other online identities, sign-in patterns, past review activity, and any odd behavior. #ai #ainews pic.twitter.com/GjK9RyMuAg

— AI ToolTracker (@AIToolTracker) June 13, 2023

“We use machine learning to look for suspicious accounts, to track the relationships between a purchasing account that’s leaving a review and someone selling that product,” said Amazon’s customer trust head, Dharmesh Mehta.

“Through a combination of both important vetting and really advanced machine learning and artificial intelligence – that’s looking at different signals or behaviors – we can stop those fake reviews before a customer ever encounters it,” he added.

Amazon has had some previous success catching sellers that use fake reviews; the e-commerce giant has found over 23,000 social media groups with over 46 million members, all of which facilitated fake reviews on e-commerce platforms.