Amazon’s shopping AI now checks whether messages are scams

When an Amazon message raises a doubt, customers can now put the question to Alexa for Shopping instead of going straight to customer service. Amazon’s shopping AI can assess whether the communication came from Amazon’s systems or is a scam, and the feature is available on the retailer’s website and mobile app.

The scale of the problem is visible in one number: 360,000 customers contact Amazon every year to ask whether a message is real. Amazon says the new system checks suspicious communications against billions of messages sent by the company worldwide, examining sender information, content, timing, message metadata and other details.

The tool follows Amazon’s earlier efforts to handle fake order confirmations, Prime membership renewal alerts, account suspension warnings, package delivery notifications and job scams. Earlier this year, the retailer introduced [email protected], where anyone can forward a suspicious message and receive an answer, alongside an online form offering the same service.

The move fits Amazon’s broader shift for Alexa toward shopping assistance and proactive help. The assistant can already find personalized deals, build shopping guides, reorder everyday products, provide AI-generated product overviews and transcribe handwritten shopping lists. Amazon has also added alerts for events such as a favorite brand launching a product, a new book from a favorite author or a new record from an artist.

In practical terms, a customer unsure about a message claiming to come from Amazon now has another route through the website or app, while email and online-form checks remain available. The limit is clear: this is an Amazon-origin check, not a general-purpose scam detector. Amazon’s claim that the AI can “definitively” confirm authenticity comes from the company itself, and its promised improvement depends on customers continuing to report suspicious messages.