AI Hyper-Personalised Phishing Impersonating Amazon
AI tools allow criminals to craft highly personalised Amazon phishing emails referencing your real recent purchases, your first name, and details scraped from social media, making fake messages nearly indistinguishable from genuine Amazon communications.
Part of: AI Hyper-Personalised Phishing Scams
Last reviewed: 8 June 2026
Traditional phishing relied on generic messages sent to millions of addresses at once. AI-assisted phishing is different: large-language models can generate personalised messages at scale, weaving in real details — your name, a product category you recently purchased, your approximate location — gathered from data-broker profiles, prior breaches, or public social media posts.
Amazon is a natural cover identity for this type of attack. The platform sends a constant stream of transactional emails — order confirmations, shipping updates, delivery notices — that customers are conditioned to open and act on. An AI-crafted fake email that references a specific product category you genuinely buy, addresses you by your real first name, and mimics Amazon's exact email template is far harder to dismiss on instinct alone.
The payload of these personalised attacks is typically credential theft or payment-card harvesting, delivered through a polished fake amazon.com sign-in page or a phone call to a fraudulent support line staffed by live agents who have been briefed on the victim's profile.
How this scam works on the Amazon brand
The scammer purchases a data-broker record containing your email, name, rough location, and sometimes purchase-interest categories. An AI tool generates a plausible Amazon email — for example, a shipping delay notice for a smart home device or a security alert about your account. The level of detail makes the message feel uniquely addressed to you rather than mass-broadcast.
The email's call to action leads to a lookalike amazon.com page. Because the email content is specific to you, you are more likely to believe the account issue is real and enter your credentials. Some campaigns follow the phishing email with an AI-generated phone call — a voice that sounds natural, using your name and referencing the 'issue' described in the email — to validate the fraudulent narrative.
The personalisation also extends to timing: AI systems can schedule delivery to coincide with peak Amazon shopping periods such as Prime Day, Black Friday, or shortly after a major news event about data security, when users are primed to act on security warnings.
Common red flags
- An Amazon email references your name and a plausible product category but still links to a non-amazon.com domain
- The message references a purchase type or brand preference that is accurate but that Amazon would not typically include in a transactional email
- You receive a follow-up phone call or text that references the same specific details mentioned in the email
- The email prompts you to click and re-enter your Amazon password or card details for an account or security reason
- The level of personalisation feels unusual for a routine Amazon transactional email — genuine Amazon emails are templated and do not include lifestyle details
- Sender email address fails close inspection — an amazon.com display name masking a non-Amazon domain
How to protect yourself
- Always navigate to amazon.com directly to check for any account alerts — never follow links in emails, no matter how personalised they appear
- Enable Amazon two-step verification so that credential theft alone is not enough to take over your account
- Review what personal data data brokers hold about you; services such as DeleteMe can help reduce your exposure
- Use a unique email address exclusively for your Amazon account so personalised phishing targeting that address stands out
- Be especially critical of highly personalised emails — more detail does not mean more legitimacy
- Use Amazon's Message Centre (amazon.com/mc) to verify whether a message about your account is genuine
How to report it
- Forward the phishing email to [email protected]
- Report it to the Anti-Phishing Working Group at [email protected]
- File a complaint with the FTC at reportfraud.ftc.gov
- If you entered credentials or payment details, change your Amazon password immediately, enable two-step verification, and contact your bank
Frequently asked questions
How does AI make phishing emails more convincing?
AI language models can generate grammatically perfect, contextually relevant text at scale, incorporating personal details from data-broker profiles. The result is an email that feels individually written rather than mass-produced.
How can I tell a personalised phishing email from a real Amazon email?
Check the actual sender domain, not just the display name. Go directly to amazon.com to verify any account alert. A real Amazon email never asks you to re-enter your password or card details through a link.
Where do scammers get my personal details for these targeted attacks?
Data broker databases, prior data breaches, social media profiles, and public purchase review histories all contribute. Reducing your data-broker exposure and using unique email addresses per service limits the data available.