Chapter 10 · AI-Powered Scams
AI Scams: When You Can't Trust What You Hear or See
Every scam in this guide relies on convincing you that a message, a call, or a person is who it claims to be. Artificial intelligence just made that far easier to fake. A stranger's voice can now sound exactly like your child's. A video call can show a face that isn't real. This chapter covers the AI tools scammers have added to the old playbook — and the one habit that still beats all of them.
Key takeaways
- AI didn't invent new scams — it made the old ones (the grandparent call, the romance scam, the CEO fraud email) far more convincing.
- Hearing a familiar voice or seeing a familiar face on a call is no longer proof of who you're talking to.
- A private “safe word” with family is now one of the most effective defenses against voice-cloning scams.
- AI-written phishing emails no longer have the typos and broken grammar that used to give scams away.
Voice cloning: the grandparent scam gets an upgrade
Chapter 4 covered the classic grandparent scam — a caller pretends to be a panicked relative in trouble. AI voice cloning removes the biggest weakness of that scam: the voice used to sound wrong. Now it doesn't have to.
With as little as a few seconds of audio — pulled from a social media video, a voicemail greeting, or a public post — scammers can generate a voice clone convincing enough to fool a parent or grandparent on the phone. The call comes in sounding exactly like your family member, crying, scared, and asking you not to tell anyone.
The defense doesn't change just because the voice is real-sounding: hang up and call that person directly on a number you already have. If they don't answer, call another family member. Never send money based on a voice alone, however certain you feel.
Deepfake video calls
Chapter 6 already warns that AI face-swapping can fake a live video call in romance scams. The same technology now shows up in business and job scams:
- Fake job interviews: a “hiring manager” on a video call is actually a real-time deepfake, used to steal your personal information during onboarding or to run a fake company that never existed.
- Fake executive video messages: an employee receives a video “from the CEO” authorizing an urgent wire transfer — the face and voice are both synthetic.
- Fake celebrity endorsements: AI-generated video of a public figure “recommending” an investment or product (see Chapter 9's crypto giveaway scams for the most common version).
A video call is a stronger signal than a phone call, but it is no longer a guarantee. If a video interaction leads to an urgent request for money, credentials, or personal information, verify independently before acting — the same rule as everything else in this guide.
AI companions and accelerated romance scams
Chapter 6's romance scams and Chapter 9's “pig butchering” scams used to require a real person typing messages for weeks to build trust. AI chatbots can now run that entire relationship-building phase automatically, at scale, in dozens of conversations at once — remembering details, matching your tone, and never running out of patience or time zones.
This doesn't change the pattern, just the volume: a stranger who won't video chat, moves the conversation off-platform, and eventually raises money or investing is still the same warning sign, whether a person or a bot is typing.
AI-written phishing and business email compromise
Chapter 3 lists a generic greeting and awkward grammar as two of the four phishing red flags. AI writing tools have quietly removed both. A phishing email can now:
- Use your actual name, pulled from a data breach or LinkedIn.
- Match the tone and formatting of your real bank, employer, or vendor perfectly.
- Reference real, specific details about your company scraped from public sources — making a fake “invoice update” from a vendor look completely legitimate.
This is especially dangerous for business email compromise, where a fake “CEO” or “vendor” email asks an employee to urgently change a payment account. The two red flags that still work: manufactured urgency, and any request to change payment details by email alone. Always verify by phone, using a number you already have on file — never one in the email.
Fake AI investment and trading tools
A newer wrinkle on Chapter 9's fake investment platforms: apps and “trading bots” marketed as AI-powered, claiming the algorithm guarantees returns no human trader could match. The AI label is used the way “guaranteed returns” always has been — to make an impossible promise sound like advanced technology instead of a lie. The same rule applies: no legitimate investment guarantees returns, artificial intelligence or not.
The one rule that still works
Rules that stop most AI scams
- Agree on a family “safe word” that isn't posted anywhere online, to use if someone calls claiming a relative is in danger.
- Treat a familiar voice or face on a call as one data point, not proof — especially if it's paired with urgency or a request for money.
- Never send money, gift cards, or crypto based on a call or video alone.
- Be extra cautious of “AI-powered” investment tools promising guaranteed or unusually high returns.
- For unexpected payment-change requests at work, always verify by phone using a number you already have — not one in the email.
Want every chapter in one place?
This guide is free to read here. If you'd like the complete book — checklists, scripts for handling a scam in progress, and every chapter offline — it's available as an eBook.