Shopping on Amazon? Then read this

Amazon announced a few days ago that it will try to help customers better understand other customers' reviews of products by introducing generative artificial intelligence. The main goal of the new functionality is to save you the "pain" of reading dozens of individual reviews. The retail giant will use the new technology to write short texts directly on the product page. These texts will focus on product features and customer opinions. This text will make it easier to understand common themes in the reviews, Amazon emphasized.

In addition to the text summary, Amazon will use artificial intelligence to highlight key product features in the form of clickable buttons. For example, if a customer wants to learn more about a product’s “ease of use” or “performance,” they can click the button to see only reviews that mention those terms.

Amazon previously offered similar functionality by highlighting frequently used words in reviews, which were also available as buttons.

The new AI-powered features will first be available to a select group of U.S. mobile shoppers for a “wide range” of products, Amazon said. During these tests, the company will work to learn and adapt its AI models to improve their performance, and it plans to gradually expand the feature to other categories as it becomes available to a wider group of customers.

Of course, AI-generated summaries will only be as good as the quality data they include. For years, Amazon has struggled with fake and misleading product reviews, including paid reviews.

In 2021, the company admitted that it had already blocked 200 million fake reviews in the previous year. They have also spent years trying to crack down on the origin of fake reviews through lawsuits and other actions, including lawsuits against sellers who bought fake reviews. Last year, Amazon filed a lawsuit against the administrators of 10,000 Facebook groups that were involved in the trade of fake reviews.

The FTC (Federal Trade Commission) also recently got involved, forcing a dietary supplement manufacturer to pay $600,000 in a case of Amazon review fraud – a case where products are grouped together in a single listing to boost the ratings of one product with the good reviews of another.

With the increasing capabilities of artificial intelligence, fake reviews can now be even harder to detect as the technology evolves to sound more human, which could lead to another explosion of fake reviews. This would mean that Amazon's AI-generated review summaries will be less useful unless the company finds other ways to prevent fake AI-written reviews.

Amazon is addressing concerns about fake reviews today, saying it will only aggregate reviews that come from verified purchases, and that it continues to “invest heavily in resources” to proactively prevent fake reviews.