Amazon is exploring an ambitious initiative to build a comprehensive database of products sold worldwide, including items that are not available on its own marketplace. The proposed project, internally described as the World Product Graph, aims to bring together product information from brands, manufacturers and competing retailers to improve product discovery, comparisons and personalised shopping recommendations. The initiative could strengthen Amazon’s position as artificial intelligence (AI) transforms how consumers search for and purchase products online.
According to a Reuters report published on November 13, 2026, an internal Amazon document outlines plans to use the database to improve the company’s AI-powered shopping assistant, Alexa for Shopping, and counter the growing influence of third-party AI shopping agents. The project could also help protect Amazon’s advertising business by keeping consumers within its shopping ecosystem rather than allowing external AI assistants to control product discovery and purchasing decisions. However, Amazon has questioned the relevance of the document, saying brainstorming materials do not necessarily represent its current plans.
Amazon’s World Product Graph: What Is the Company Planning?
The proposed World Product Graph is designed to create a large, structured database containing information about products sold across the global retail market.
Unlike Amazon’s existing product catalogue, which primarily focuses on products available through its marketplace, the proposed database would also include items sold by other retailers and independent brands.
The system would combine product information supplied directly by manufacturers with information collected from other retailers’ websites.
This could include:
- Product names and descriptions.
- Technical specifications and product attributes.
- Available colours, sizes and configurations.
- Product images and other listing information.
- Brand and manufacturer details.
- Information useful for comparing competing products.
The objective is to make product information more comprehensive and accessible to AI-powered shopping systems.
For example, a customer searching for a particular television could receive comparisons involving products sold on Amazon and competing retail websites. Similarly, someone looking for running shoes could compare specifications and product features across multiple brands without manually visiting several online stores.
The proposed database could therefore help Amazon become a broader product discovery destination rather than a marketplace limited to its own inventory.
Why Amazon Wants to Own More Product Data
Product information has become increasingly important as consumers move from traditional keyword-based searches towards conversational AI assistants.
Previously, shoppers typically searched for products using retailer websites or search engines such as Google. They would browse listings, compare prices, read reviews and visit different websites before making a purchase.
AI shopping assistants are changing this process.
Consumers can now describe what they need in ordinary language and ask an AI system to research available options, compare specifications and recommend products.
This creates a challenge for Amazon because the company may lose control over the initial product discovery process if customers increasingly rely on independent AI platforms.
The World Product Graph could help address this challenge by giving Amazon access to a wider collection of structured product information.
A more comprehensive database could allow its AI assistant to answer shopping questions more accurately, recommend relevant products and present comparisons that extend beyond Amazon’s own marketplace.
It could also make Amazon more competitive against AI-powered shopping experiences developed by Google, OpenAI, Meta and other technology companies.
Amazon Wants to Improve Alexa for Shopping
One of the proposed project’s major objectives is to strengthen Amazon’s AI shopping assistant, previously known as Rufus and renamed Alexa for Shopping in May 2026.
The assistant is designed to help customers research products, compare options, receive personalised recommendations and complete purchases.
Amazon has been investing in conversational shopping technology to make product discovery easier.
However, the internal document cited by Reuters reportedly acknowledged that the assistant sometimes provides inaccurate or incomplete responses and can be expensive to operate.
Amazon disputed the document’s characterisation, saying its assistant is approximately 97% accurate.
The proposed World Product Graph could help improve the system by providing more structured and comprehensive information about products.
A larger database could help Alexa for Shopping:
- Improve product recommendations: Understand product specifications and customer requirements more effectively.
- Compare competing products: Provide information about alternatives available across different retailers.
- Answer detailed questions: Respond to specific queries about features, compatibility and product differences.
- Improve personalisation: Match recommendations with individual shopping preferences.
- Support purchase decisions: Help customers identify suitable products and complete transactions.
Amazon’s broader strategy is to make its AI assistant a central part of the shopping experience, reducing the need for customers to visit multiple platforms.
Coca-Cola and Bayer Reportedly Among Companies Involved
The development of a global product database requires cooperation from brands and manufacturers because they possess detailed information about their products.
According to Reuters, companies including Coca-Cola and Bayer have reportedly agreed to provide product information for the initiative.
Amazon also explored a potential trial with consumer goods company Procter & Gamble.
The participation of major manufacturers would be important because product information collected from retailer websites may not always be complete or consistent.
Direct information from manufacturers could help Amazon build a more reliable database containing specifications, descriptions and other product details.
For consumer brands, participating in such a system could provide additional visibility across digital shopping channels.
However, the availability of accurate product information and the ability to update it regularly will remain important challenges, particularly when products have multiple versions, regional specifications or changing availability.
AI Shopping Agents Are Becoming a Threat to Amazon’s Business Model
The proposed initiative comes as independent AI shopping agents begin challenging the traditional online retail model.
AI agents can research products, compare prices across retailers and potentially complete purchases on behalf of consumers.
Instead of opening Amazon and searching for a product, a customer could ask an external AI assistant to find the best available option across several shopping platforms.
This could weaken Amazon’s influence over product discovery and reduce the amount of customer activity taking place directly on its website and application.
Amazon has already taken steps to restrict certain external shopping agents from accessing its marketplace.
The company blocked Meta’s Muse shopping agent from its website, while its legal dispute with Perplexity over the Comet shopping agent has highlighted disagreements over how third-party AI tools should access retail platforms.
The debate is becoming increasingly important as technology companies compete to control the relationship between consumers, product information and online transactions.
Why AI Shopping Agents Could Threaten Amazon’s Advertising Revenue
Amazon’s advertising business is another major reason behind its interest in controlling product discovery.
The company operates one of the world’s largest retail media advertising businesses, allowing brands and sellers to pay for sponsored product placements and promotional visibility.
When customers search for products on Amazon, sponsored listings can influence which items they see and purchase.
However, independent AI shopping agents could change this relationship.
If consumers ask an external AI assistant to recommend the best product, the AI system may select an item based on price, specifications and customer requirements rather than sponsored placements on Amazon.
This could reduce the influence of Amazon’s advertising ecosystem if more purchasing decisions begin outside its platform.
A Reuters analysis published earlier in November highlighted how AI-led shopping could challenge the advertising businesses of Amazon and Walmart.
The report noted that the two retailers together account for a dominant share of the US retail media advertising market.
For Amazon, therefore, controlling product information is not just about improving shopping recommendations. It is also about maintaining its position in online advertising and customer acquisition.
Amazon’s World Product Graph Could Launch in 2027
The internal document reviewed by Reuters reportedly included an aspirational launch date of October 2027 for the World Product Graph initiative.
However, the timeline remains uncertain.
Reuters could not independently confirm when the project would launch or whether the proposed schedule remains active.
Amazon also questioned the document’s relevance, emphasising that internal brainstorming materials do not necessarily represent the company’s current strategy.
The distinction is important because the World Product Graph should be understood as a reported initiative under development rather than a confirmed product launch.
If implemented at scale, the database could become an important part of Amazon’s broader AI strategy.
It would also require substantial work involving data collection, product matching, verification, updates and coordination with manufacturers and retailers.
The Bigger Picture
Amazon’s proposed World Product Graph highlights a major shift in the competition between traditional e-commerce companies and AI-powered shopping platforms. As consumers increasingly use conversational assistants to research and compare products, control over product information could become as important as control over inventory, logistics and payment systems. Amazon wants to strengthen its position by creating a comprehensive database that helps its AI assistant deliver better recommendations while potentially reducing customers’ dependence on competing platforms.
The initiative also reflects a broader commercial battle over the future of online advertising. If independent AI agents become the primary way consumers discover products, retailers could lose some control over sponsored placements and customer relationships. Amazon’s strategy suggests that the company wants to remain central to shopping decisions even as the technology used to make those decisions changes.
Looking Ahead
Amazon’s next challenge will be turning the proposed World Product Graph into a reliable and scalable product information system. Cooperation from manufacturers, accurate information from external retailers and regular database updates will be essential. The company’s ability to improve Alexa for Shopping while maintaining customer trust will also determine whether the initiative can become a meaningful competitive advantage.
The broader retail industry will be watching how Amazon responds to the growing popularity of independent AI shopping agents. If consumers increasingly delegate product research and purchasing decisions to AI, competition could shift away from traditional search rankings and marketplace visibility towards data quality, recommendation accuracy and access to real-time product information. Amazon’s proposed initiative could be an important step in preparing for that future, although its timing and final implementation remain unconfirmed.
Get the day’s top stories in your inbox
One concise email. No spam, unsubscribe anytime.
