Illustration of Amazon Unveils AI Shopping Assistant Rufus for U.S. Customers

Amazon Unveils AI Shopping Assistant Rufus for U.S. Customers

After extensive testing, Amazon has officially launched its generative artificial intelligence shopping assistant, Rufus, for all U.S. customers today.

The conversational shopping assistant is designed to enhance the shopping experience by saving time and helping customers make more informed purchase decisions, according to Amazon. Rufus is now available in the Amazon shopping app just in time for Prime Day, occurring on July 16 and 17.

Announced in January, Rufus is capable of answering specific questions about products, such as their ease of maintenance and material composition. The AI-powered assistant can also provide product recommendations, comparisons, and updates. Additionally, customers can track packages, review past orders, and even seek advice on topics unrelated to shopping, such as what ingredients are needed for a soufflé or how to plan a summer party.

Amazon, the largest cloud provider, has introduced its own AI training and inferencing chips, as well as a platform called Bedrock for developers to build generative AI applications on its Amazon Web Services cloud service. Despite these advances, Amazon has not concentrated as much on developing AI products compared to competitors like Google and Microsoft.

Recently, it was reported that Amazon is working on an AI chatbot, internally named “Metis,” intended to rival OpenAI’s ChatGPT. This chatbot, powered by one of the company’s internal AI models, Olympus, will be accessible through a web browser. According to Business Insider, which cited unnamed sources and an internal document, Olympus is reportedly more powerful than Amazon’s existing public AI model, Titan.

In March, Amazon completed a $4 billion investment in AI startup Anthropic, marking its largest investment in an external company to date. Anthropic, which uses AWS as its primary cloud provider, is set to utilize Amazon’s AI chips to build, train, and deploy its future models.

Popular Categories


Search the website