9 Common AI Scams and How Each One Works
AI scams are frauds that use artificial intelligence to make an old trick more convincing or cheaper to run at scale. The most common ones are voice-clone emergency calls, deepfake video calls from "executives", polished AI-written phishing, romance scams run with generated photos and chat, fake AI investment platforms, counterfeit AI apps, fake job offers, AI-generated shop and review sites, and sextortion using fabricated images. The technology is new, but the goal is the same as always: get you to send money or hand over access before you check.
1. The voice-clone emergency call
How it works: A scammer collects a few seconds of someone's voice from social media videos or voicemail greetings, then uses a cloning tool to generate new speech. You get a call that sounds like your child, parent or partner, panicked and asking for money: an accident, an arrest, a stranded trip.
The pattern: Emotional shock, urgency, secrecy ("don't tell Mum"), and a payment method that is hard to reverse.
Defense: Hang up and call the person on their usual number. Agree a family code word in advance.
2. The deepfake executive
How it works: An employee in finance receives a message, then a video or voice call, apparently from a senior leader asking for an urgent confidential transfer. The face and voice are synthesized or manipulated.
The pattern: Authority, secrecy, a deadline, and a payment that bypasses normal approval.
Defense: Every payment request, however senior the sender, goes through standard approval and a call-back on a known number.
3. AI-polished phishing
How it works: Phishing emails and texts used to be easy to spot by their errors. Now scammers use language models to write fluent, on-brand messages in any language, often personalized with your name, employer or recent purchases.
The pattern: A link to "verify", "unlock" or "confirm" something, leading to a login page that harvests your credentials.
Defense: Do not log in through links in messages. Open the app or type the address yourself.
4. Generated romance
How it works: A profile with attractive, AI-generated photos strikes up a relationship. Chat may be partly automated, letting one scammer run many conversations. Over weeks, trust builds, then come the requests: a medical bill, a visa fee, a "sure thing" investment.
The pattern: Fast intimacy, reasons they can never meet or video chat properly, and eventually money.
Defense: Run a reverse image search on photos, insist on a live video call, and never send money to someone you have not met in person.
5. Fake AI investment platforms
How it works: Ads or social posts promote an "AI trading bot" or "AI-powered fund" promising steady, high returns. Victims see a dashboard showing gains, are encouraged to invest more, then find they cannot withdraw. Some use deepfake videos of celebrities endorsing the scheme.
The pattern: Guaranteed or unusually high returns, pressure to recruit others, and fees required before withdrawal.
Defense: Check whether the firm is registered with your country's financial regulator. Treat celebrity endorsements in ads as suspect.
6. Counterfeit AI apps and tools
How it works: Fake apps or browser extensions pose as popular AI chatbots or image generators. They may charge subscription fees for nothing, show endless ads, or install malware that steals passwords.
The pattern: Downloads from unofficial sources, requests for broad permissions, payment up front.
Defense: Download only from official app stores or the developer's own site, check the publisher name, and read the permissions.
7. Fake job offers
How it works: AI makes it easy to produce professional job listings, offer letters and even automated "interviews" by chat. The victim is asked to pay for training or equipment, or to provide bank details and ID documents for "payroll."
The pattern: A job offer without a real interview, upfront fees, or requests for sensitive documents early on.
Defense: Look up the company independently and contact its official HR. Real employers do not charge you to start working.
8. AI-generated shops and reviews
How it works: Scammers spin up convincing online stores with generated product photos and descriptions, backed by batches of fake reviews. You pay, and nothing arrives, or something worthless does.
The pattern: Prices far below market, a new domain, no physical address, and payment only by bank transfer or crypto.
Defense: Search the shop name with "scam" or "reviews" on independent sites, pay by credit card where possible for buyer protection, and check how long the domain has existed.
9. Fabricated-image sextortion
How it works: Scammers take ordinary photos from someone's social media and use AI to create fake explicit images, then threaten to share them unless paid. Teenagers are often targeted.
The pattern: Threats, short deadlines and demands for payment.
Defense: Do not pay, as payment often leads to more demands. Keep evidence, report to the platform and to police, and talk to someone you trust. Many countries have services that help remove such images.
Why AI makes scams more effective
AI lowers three costs for criminals. Language is no longer a barrier, since a scammer can write natively in any language. Personalization is cheap, because details scraped from public profiles can be folded into each message automatically. Volume goes up, because one person can run hundreds of conversations at once. None of this changes the core of the scam, which is why the same defenses keep working.
The defenses that work against all of them
- Slow down. Urgency is the scammer's main tool.
- Verify through a second channel you already trust.
- Never share codes or passwords with someone who contacted you.
- Be wary of irreversible payments: gift cards, crypto, wire transfers.
- Limit what you share publicly, including voice and video clips if you are concerned about cloning.
- Use two-factor authentication on important accounts.
Text tools can play a small role. If a message seems off, an AI detector can estimate whether it was machine-written, though that tells you nothing certain about whether it is a scam, since many genuine companies use AI for email. And a text summarizer can help you digest a long, confusing contract or terms page so you can spot unusual clauses before you sign up for anything.
FAQ
What is the most common AI scam?
AI-polished phishing is likely the most widespread because it is cheap to send in bulk. Voice-clone emergency calls and deepfake payment requests get more attention because individual losses can be large.
Can scammers clone my voice from a short clip?
Voice cloning tools can work from fairly short samples, though quality varies. A family code word and a call-back rule protect you regardless of how good the clone is.
Are older people the only targets?
No. Scams target every age group. Job scams and sextortion often target younger people, investment and romance scams target a wide range of adults.
Where should I report an AI scam?
Report to your bank if money is involved, to the platform where contact happened, and to your country's fraud reporting agency or police.