Amazon seeks to amass data on every item for sale worldwide to fight AI shopping agents

Amazon is developing a ambitious plan to gather data on virtually every item for sale worldwide to enhance its shopping experience and compete against rivals, according to an internal document reviewed by Reuters. The project, called World Product Graph, aims to collect product information from virtually every retailer, even those Amazon doesn't stock, in order to become the "everything store" and rival Google's dominance in online search.

The data will also help combat the rise of AI shopping agents, as Amazon's own Rufus search engine often provides inaccurate information. Several major brands, including Coca-Cola and Bayer, have already agreed to contribute product data to the project. Amazon's internal staff is being reorganized to work on the initiative, which could be implemented by October 1, 2027.

While Amazon believes Alexa for Shopping is currently 97% accurate, the company is still seeking to reduce costs and improve the quality of its AI-driven search. The World Product Graph could also protect Amazon from the growing threat of alternative AI shopping agents, such as Meta's Muse, which has already gained a significant following in the US and Canada.