AI Scam Pattern Library 2026
A free, curated catalogue of how AI-generated scams actually appear in the wild — with redacted real examples, detection signals, and recovery steps.
Generative AI has made it easy to mass-produce convincing scam content — phishing emails that read like a CEO wrote them, voice calls that sound exactly like a grandchild in distress, fake LinkedIn recruiters whose profile photos and work history pass casual inspection. The defensive challenge has changed: the surface-level "spot the typo" advice no longer works because AI doesn’t make typos.
This library catalogues each major pattern with what it actually looks like, the specific signals that still distinguish AI-generated from human content, and what to do if you’ve been targeted. We update it as new patterns appear. Report a new pattern if you’ve seen something we should add.
AI-Cloned Voice Calls Targeting Elderly Parents
The grandchild-in-trouble call: AI voice cloning makes the synthetic voice indistinguishable from a real family member. India-heavy due to overseas-child demographics; same pattern globally.
ChatGPT-Written CEO Impersonation Emails (BEC)
The Business Email Compromise variant where AI removes the traditional tells — perfect grammar, executive tone, plausible context. Six signals that survive.
AI-Generated Romance Scam Messages (Pig-Butchering)
How AI chatbots are now running the months-long emotional grooming half of pig-butchering scams. The conversational signals that still distinguish AI from a real human partner.
Scan any link, file, or message free
If you got a suspicious message, run it through ScanTotal’s URL scanner or SMS analyser before responding.
Open Scanner AI App Safety Check