Amazon wants an everyday text message to become the beginning of a shopping trip. Alexa for Shopping can now receive requests by SMS from a limited group of invited US customers, look for products, compare options and help complete orders. Announced on 6 November 2026, the feature extends Amazon’s AI assistant beyond the app and website. It is not a universal launch and shoppers outside the initial invitation programme should not assume the text service is already available to them.
The attraction is obvious when the request is mundane. A shopper might remember they need dog food while walking home, send a message and return to something else. Amazon says the assistant can handle repeat purchases, look for a birthday present, work from a photograph or watch for a target price. Instead of sending a link to a search page, it can prepare a basket and, with the customer’s approval where required, move towards purchase.
What Alexa for Shopping can do by text
Amazon describes a conversation that can move between text messages, its shopping app, the website and compatible Alexa devices. A person could start looking for a gift on the train, inspect the choices properly later and continue from the same task. Preferences supplied by the shopper, such as preferred product types, are intended to carry over rather than having to be repeated. That continuity is central to the pitch: the assistant remembers what is being discussed and the action the customer wanted to take.
The service goes beyond basic search. It can use shopping history to help with reorders and compare prices and reviews. Amazon also describes sending a picture to request similar products. In a separate guide to its shopping assistant, the company explains features such as checking price history and automatic purchasing at a specified target price. Those capabilities have different implications from merely asking which television has the better screen.
The moment a helpful suggestion becomes a real order
There is an important dividing line between the assistant assembling a basket and spending money. Amazon says Alexa for Shopping can place purchases on a customer’s behalf, while its examples include explicit approval for a reorder. Customers should check the approval arrangements and automation settings available to their account before giving broad instructions. A message sent in a hurry may omit a brand, size or quantity that matters more than the assistant realises.
Price, delivery date and substitutions also deserve attention. An assistant can rank products according to information available to it, but that does not make every recommendation the best value for a particular person. A discounted product with the wrong specification is not a saving. The sensible approach is to inspect the final selection for important purchases and to be careful with open-ended commands such as buying whatever looks suitable.
Why this is more than another Alexa feature
The usual shopping model asks the customer to browse a catalogue. Agentic shopping reverses part of that process: the customer describes a goal and the software performs steps towards it. LiveAIWire has already examined Alexa's automatic buying and price alerts and Amazon’s work with merchants beyond its own store. The text option changes where that activity starts. A shopper no longer needs to open the storefront to initiate the task.
It also shifts commercial power towards the assistant’s shortlist. Retailers will care which products appear in a conversational recommendation, and shoppers will need to know how results are selected. A familiar message thread might make a choice feel informal, but the transaction is still a commercial decision with consequences for returns, delivery and household budgets. Convenience can reduce friction in useful ways while also making it easier to purchase without giving alternatives much attention.
Who can use the text service today?
Amazon says the service began with an invitation-only group of US customers and is scheduled to expand over the following weeks. Users need to accept terms and link a telephone number. The announcement does not establish that the feature is available internationally, or that every device and product category works identically. Invitations, rollout dates and individual account settings determine what someone can actually do.
For those with access, the first test should be a small, familiar request with no costly consequences. A basic reorder or price enquiry makes it easier to see how the assistant interprets instructions and how consent works at checkout. The ambitious future is a shopping agent that handles routine errands quietly. The practical test is simpler: did it find the right product, explain the trade-offs and ask before doing anything the customer would regret?
The problem with a recommendation that feels personal
A text conversation creates an impression of individual service. The assistant can refer to earlier preferences and present a short list rather than a catalogue of competing results. That can genuinely help someone who knows what they need but dislikes searching through similar products. Yet a personal recommendation is not the same as independent advice. The retailer operates the service and has commercial relationships with sellers, so customers should still consider price, delivery, quality and alternative offers.
Shoppers may also want to distinguish a reminder from permission to purchase. Asking an assistant to remember that washing powder is running low should not automatically mean authorising whatever product it chooses at whatever price is available. The ability to set a clear limit and confirm an order is part of making automation trustworthy.
Messages bring their own security questions. A customer needs to be certain they are communicating with the authentic service and understand what happens if the linked phone number changes hands. Amazon’s invitation system and account settings matter here. No sensible consumer should give payment credentials or one-time security codes to an unexpected text claiming to be an AI shopping assistant.
About the Author
Stuart Kerr is Technology Correspondent at LiveAIWire, covering artificial intelligence, cybersecurity and the social impact of emerging technology. LiveAIWire is an independent, human-led technology publication using AI-assisted research, editorial production and original AI-assisted editorial illustrations under his direction.