Scam messages, sites, and personas produced at scale by generative AI, often free of the old spelling giveaways.
Generative AI lets scammers write flawless, personalized messages in any language, spin up convincing fake websites, and run tireless chatbot “recruiters.” The old advice to spot scams by bad grammar is now unreliable.
Because content is cheap to generate, the same operation can target far more people with far more tailored lures.
Red flags
- Unusually polished, personalized outreach from an unverified source
- A “recruiter” who replies instantly at all hours
- Perfect messaging paired with a suspicious payment or link ask
Frequently asked questions
Can I still spot scams by looking for bad grammar?
Not reliably. AI now writes clean, natural messages, so grammar is a weak signal. Focus instead on what's being asked — payment, credentials, moving off-platform — and verify the source independently.