How Amazon’s AI Technology fights fake reviews

Role of AI in review moderation

When a customer submits a review, Amazon's AI systems immediately get to work. The review is analyzed for any indicators that suggest it might be fake. According to Amazon, the majority of reviews pass these authenticity checks and are posted almost instantly. However, if the AI flags a review as potentially fake, Amazon takes swift action, including blocking the review, removing it, or even taking legal steps against the involved parties. In 2023 alone, Amazon proactively blocked more than 250 million fake reviews from being posted.

Josh Meek, Senior Data Science Manager at Amazon’s Fraud Abuse and Prevention team, emphasizes the importance of this rigorous process. “Fake reviews intentionally mislead customers by providing information that is not impartial, authentic, or intended for that product or service,” says Meek. This sentiment underscores the dual responsibility Amazon carries: protecting customers from misleading information and ensuring that brands and businesses on its platform are represented fairly.

Advanced AI techniques

Amazon employs a variety of AI techniques to combat fake reviews. These include machine learning (ML) models that analyze a wealth of proprietary data such as review history, customer behavior, and reports of abuse. Natural language processing (NLP) and large language models (LLMs) further help in detecting anomalies that might suggest a review is incentivized or manipulated. Additionally, Amazon uses deep graph neural networks (GNNs) to examine complex relationships and behavior patterns, which aids in identifying and removing groups of bad actors.

Despite the sophistication of these technologies, the challenge of differentiating between genuine and fake reviews remains. Meek explains that certain factors, like rapid review accumulation due to advertising or unusual grammatical patterns, can make it difficult to assess a review’s authenticity without deeper insights, which Amazon’s proprietary systems provide.

Commitment

Maintaining a trustworthy shopping experience is Amazon's top priority. Rebecca Mond, Head of External Relations, Trustworthy Reviews at Amazon, affirms the company's dedication to this mission. “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,” says Mond. Through ongoing innovation and the use of cutting-edge AI, Amazon strives to keep its platform reliable and trustworthy for all users.