Amazon uses machine learning and graph neural networks to run billions of daily scans of its marketplace — automatically detecting fraudulent seller accounts, counterfeit product listings, fake reviews, and other forms of abuse before they reach customers.

Amazon invested more than $1 billion in 2024 alone on brand protection, employing machine learning scientists and human investigators. The AI systems scan product listings and customer interactions weekly for signals of abuse. Amazon claims its proactive AI stops "at least 99% of fraudulent listings before a brand brings it to our attention." In 2024, Amazon identified and disposed of over 15 million counterfeit products worldwide. Amazon notes that counterfeiters are increasingly adopting generative AI themselves to create more convincing fake listings and reviews, making this an ongoing arms race. Its Counterfeit Crimes Unit has pursued over 24,000 bad actors since its 2020 launch.