Key Takeaways
Consumers rely heavily on product reviews to gauge whether a beauty product will suit their specific skin or hair type, but the growing use of AI-generated reviews threatens to undermine that trust. A recent study uncovered signs of AI-generated product reviews in Amazon’s beauty category.
Conducted by Pangram Labs, an organization dedicated to helping AI have a positive impact on society, the large-scale analysis revealed the widespread use of AI-generated reviews across Amazon’s top-selling products, including electronics, home goods, and beauty products.
How Did They Discover This?
Pangram Labs utilized its market-leading AI detector tool to conduct a comprehensive analysis of nearly 30,000 Amazon reviews. These reviews came from 500 of Amazon’s top-selling products across a diverse array of categories, including but not limited to beauty, electronics, home goods, and personal care. The main goal of the analysis was to assess the authenticity of consumer feedback and identify patterns in review content that might suggest the use of AI-generated text. By using advanced language detection algorithms, Pangram Labs was able to offer critical insights into the integrity of online product reviews, helping consumers and brands alike navigate the increasingly complex digital marketplace.
Important Findings
E-Commerce Integrity at Risk as AI Reviews Infiltrate Amazon’s Beauty Category
AI reviews could potentially get out of control and threaten the trust that online shoppers have in online retailers.
“The proliferation of AI-generated reviews has the potential to break trust in the customer review system once and for all," warned Max Spero, CEO at Pangram Labs. "Amazon was the first online retailer to introduce a review feature, but the system is now threatened by large language models (LLMs) like ChatGPT, Claude, and Gemini. And this could be deeply damaging for the e-commerce giant and the wider industry.”
Although Amazon originally used generative AI to help purchasers, having a notable amount of AI-generated reviews on popular products in the beauty sector of Amazon can potentially mislead consumers into purchasing products because they assume there are people behind these five-star reviews stamped with the title "verified purchase." This can create mistrust between consumers and online retailers.
In response, Amazon is taking a stand against these reviews. Last year, Amazon filed a lawsuit against Bigboostup.com, a website that facilitates fake product and business reviews targeting Amazon. The website sold fake product reviews to bad actors to publish on their product listing pages in Amazon’s store. In a press release announcing the lawsuit, the company said it has "zero tolerance for fake reviews."
Amazon is also utilizing AI technology to verify the authenticity of reviews before they are posted. Large language models are used in conjunction with natural language processing techniques to analyze anomalies in this data that may indicate a review is fake or incentivized with a gift card, free product, or some other form of reimbursement. Amazon utilizes deep graph neural networks to analyze and understand complex relationships and behavior patterns, helping to detect and remove groups of bad actors or identify suspicious activity for investigation.
“Maintaining a trustworthy shopping experience is our top priority,” said Rebecca Mond, head of External Relations, Trustworthy Reviews, at Amazon, in a press release. “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.”
