Amazon Shifts AI Strategy to Focus on Operational Gains and Retail Rivalry

Amazon is scaling back several of its homegrown artificial intelligence models to prioritize a smaller cluster of advanced projects, according to a report from PYMNTS. The company is shifting resources away from specific products within its Nova family—including the Premier and Omni models, as well as Reel and Canvas—with some of these initiatives reportedly moving into maintenance mode.

Prioritizing Capital Discipline Over Frontier Models

The move marks a strategic pivot. Instead of competing solely for the most celebrated general-purpose frontier model, Amazon is redirecting engineering resources to strengthen its core pillars: cloud computing, retail operations, advertising, and logistics.

It is a play for economic application. By focusing on its proprietary commercial systems, Amazon aims to maximize the utility of its existing infrastructure.

The AWS Infrastructure Advantage

Amazon wins when enterprises rent its computing power or deploy models through Amazon Web Services (AWS). This is particularly evident with Amazon Bedrock. The environment gains value as organizations use it to deploy, govern, and switch between multiple AI models, regardless of whether those models were built by Meta, Anthropic, or Amazon.

A Divergent Path from Walmart

This restructuring puts Amazon in direct competition with Walmart, though the two retail giants are playing different games. Walmart has avoided the frontier-model race entirely, instead applying AI to product search, supply chains, employee productivity, and store operations.

The battle has shifted from model ownership to operational context. Amazon’s edges are its fulfillment centers, marketplace sellers, advertising inventory, Prime memberships, and AWS. Walmart’s strengths lie in physical stores, local inventory, supplier relationships, employee workflows, and purchase data.

The Rise of the AI-Powered Shopper

This strategic shift arrives as consumer behavior evolves. Data from PYMNTS Intelligence and Visa Acceptance Solutions, detailed in the report Global Digital Shopping Index: The AI-Powered Shopper Has Arrived, shows that 53% of purchases are now made via mobile devices.

The most significant change is happening before the checkout. Shoppers are increasingly using smartphones as in-store tools to discover products, compare prices, read reviews, check inventory, access loyalty offers, and complete payments.

Logistics Over Popularity

By narrowing its AI ambitions, Amazon is positioning itself to win the AI economy by lowering the cost of fulfilling orders and increasing the volume of ads and inventory movement. It is a gamble on practical advantages in logistics and cloud infrastructure over the popularity of any single AI model.