Amazon Explores Global Product Database to Counter Google and Power AI

Amazon is reportedly developing the World Product Graph, an ambitious database cataloging global online merchandise to reduce reliance on Google and power its AI shopping assistants.

Amazon Explores Global Product Database to Counter Google and Power AI
Photo: Unsplash / Rubaitul Azad

Amazon is reportedly exploring an ambitious initiative to build a comprehensive database containing detailed information on every product sold across the internet, including items not offered on its own e-commerce platform. According to an internal document reviewed by Reuters, the project, dubbed "World Product Graph," aims to address two strategic challenges facing the company in the artificial intelligence era.

The project is designed to prevent human shoppers from leaving Amazon to research products and reviews on Google or Reddit, while also ensuring that AI shopping agents rely directly on Amazon's data repository. By securing this foundational infrastructure, the company intends to capture transactions even as consumer purchasing shifts toward automated AI bots.

Addressing Internal AI Limitations

This strategic push aligns with recent measures implemented by Amazon, including the launch of the "Shop Direct" feature last year, which displays direct purchasing links for external retailers. In recent months, the company has also blocked access by competing AI shopping agents, such as Meta's Muse bot.

However, the internal document reveals that the World Product Graph is also intended to overhaul Amazon's proprietary AI shopping assistant, Alexa for Shopping, formerly known as Rufus. The report notes that the document candidly admits that many of Rufus's current answers are factually incorrect or incomplete, and that running queries through the AI engine entails exceptionally high operational costs.

Engineers hope that feeding the system structured data covering the entire retail market will improve the assistant's accuracy, allowing the company to deploy smaller and more cost-effective language models.

Implementation Challenges and Company Stance

Despite an ambitious target launch date of October 1, 2027, outlined in the document, Reuters emphasizes that it remains uncertain whether or when the initiative will be fully realized.

Responding to the reports, an Amazon spokesperson stated that outdated drafts and internal documents often present an incomplete picture, noting that the existence of a brainstorming document does not necessarily reflect the company's current roadmap.

Related News