Amazon is pointing its AI at a problem that’s been bleeding consumer trust for years: the flood of fake messages impersonating the company. The retailer says roughly 360,000 customers annually contact support just to ask whether an email, text, or call claiming to be from Amazon is legitimate. Starting now, they can skip the phone queue and ask Alexa for Shopping instead.
The AI works by checking the suspicious message against the billions of communications Amazon has actually sent worldwide. It examines sender information, content, timing, and metadata before issuing what Amazon calls a “definitively” verdict on authenticity. The system also learns from user reports, which means each forwarded scam message theoretically makes the next detection faster.
This builds on earlier anti-fraud efforts. Amazon already operates verify@amazon.com, where anyone can forward a suspicious email and get a confirmation response, plus an online form on its customer service site that does the same job. The scam landscape it’s fighting includes fake order confirmations, Prime renewal alerts, account suspension warnings, package delivery notices, and job offers — the full menu of social engineering tricks.
The timing says something about Amazon’s broader strategy. Alexa has been repositioned hard toward shopping assistance this year, with personalized deals, shopping guides, reorder help, and AI product overviews already live. On Tuesday the company added alerts for things like new product drops from favorite brands or new books from followed authors. Scam verification slots neatly into that same utility-first framing — less futuristic assistant, more practical shopping companion. The question is whether consumers will trust an Amazon AI to judge whether something is an Amazon scam. That’s a conflict of interest worth watching, even if the underlying verification logic is sound.
