Amazon Introduces Automated Review Summaries Powered by AI
Today, Amazon unveiled a fresh generative AI capability that condenses product reviews. Initially accessible to a limited number of mobile shoppers in the U.S., covering a wide range of products, this AI tool generates a concise summary paragraph that captures prevalent themes from customer reviews. In June, the company confirmed its experimentation with an AI-driven summarization tool, and now it commences its official launch. CEO Andy Jassy emphasized earlier this month that AI is a fundamental aspect of Amazon’s operations.
The idea behind the ML-generated summary is to allow buyers to get an idea of the impressions of their peers without having to do multiple reviews manually. The summary includes a short paragraph describing the customer consensus: It’s kind of like an AI-powered version of the “Critics’ Consensus” and “Public Says” reports found on Rotten Tomatoes. “Customers like the stability, ease of use and performance of a digital device,” reads a sample summary shared by Amazon. “They mention that it’s much faster, the picture/streaming speed is excellent and it’s a simple device to connect. They also appreciate the performance and say that it works as expected and works great with the LG 3D Smart TV.
The summary is followed by clickable tags that introduce relevant themes and common words from customer reviews. (It’s similar to the existing keyword feature in company reviews.) Clicking on one will take you to full reviews on the selected theme.
The elephant in the room is Amazon’s reputation for fake reviews. While the retailer says it “proactively blocked more than 200 million suspected fake reviews” in 2022 alone — and is known to sue the culprits (and in extreme cases get help from the FTC), that hardly means the company will spot and block them all. There is also the question of whether AI-powered fake reviews (using ChatGPT or similar tools) are more challenging for Amazon to detect than those written by humans.
The company’s strategy includes only releasing the summarization tool for verified purchases using AI models that it claims can detect natural reviews — and calling in human researchers when necessary. “We continue to invest significant resources in preventing fake reviews,” said Vaughn Schermerhorn, director of Amazon Community Shopping. “This includes machine learning models that analyze thousands of data points to detect risk, including relationships with other accounts, login events, review history and other signs of unusual activity, as well as expert researchers who use advanced fraud detection tools to analyze and prevent fake reviews from appearing on our store. The new AI-generated review highlights only use trusted to our collection of reviews from our Verified Purchases, ensuring that customers can easily understand the community’s opinions at a glance.
