Artificial intelligence can write convincing emails, clone voices, create realistic images, and generate videos that appear authentic. These tools offer real benefits, but scammers also use them to impersonate relatives, recruiters, banks, government agencies, executives, and trusted companies.
The best defense is not learning to spot every deepfake. AI-generated content will keep improving. Instead, build a repeatable verification habit: pause, leave the conversation, and confirm the request through a trusted channel.
What is an AI scam?
An AI scam uses artificial intelligence to make traditional fraud more convincing or easier to distribute. The underlying tricks are familiar, but AI helps criminals create polished messages, believable identities, cloned voices, fake videos, and personalized phishing campaigns quickly.
Common AI-enabled scams include:
- Voice calls that imitate relatives or managers
- Deepfake videos promoting fake investments
- Personalized phishing emails
- Fake recruiter profiles and job interviews
- AI-generated social media accounts
- Fabricated identification documents
- Tech-support chatbots
- Fake invoices and payment requests
- Fraudulent product reviews and websites
The FBI’s 2025 Internet Crime Report recorded more than one million suspected internet-crime complaints and over $20 billion in reported losses.
How serious are AI scams?
AI fraud now causes measurable financial harm. According to the FBI’s April 6, 2026 report on cryptocurrency and AI scams, complaints involving artificial intelligence accounted for 22,364 reports and nearly $893 million in losses during 2025.
The FBI says scammers use fake profiles, voice clones, fabricated identification, and believable videos depicting loved ones or recognizable people. These tools can make a scam look professional, personal, and urgent even when the criminal knows very little about the target.
What warning signs reveal an AI scam?
Visual glitches and robotic voices may provide clues, but do not depend on them. Modern AI content can look and sound convincing.
Focus on the behavior behind the request:
- Pressure to act immediately
- Instructions to keep the situation secret
- Requests for gift cards, cryptocurrency, cash, or wire transfers
- An unexpected change in payment information
- Requests for passwords or verification codes
- Refusal to communicate through normal channels
- A recruiter asking for sensitive data before an interview
- A relative calling from an unfamiliar number
- A bank representative asking you to move money for “protection”
- A public figure promising guaranteed investment returns
CISA’s phishing prevention guidance recommends verifying suspicious requests through a known contact method rather than replying or using a link or number supplied in the message.
How can you verify a cloned voice or deepfake call?
Never trust a voice simply because it sounds familiar. The FTC’s voice-cloning consumer guidance recommends calling the person back using a phone number you already know instead of trusting the incoming call.
Use this verification process:
- End the conversation
Do not let urgency keep you on the call. - Contact the person independently
Use a saved number, official company app, or verified website. - Ask a private question
Create a family safe word or ask something a stranger could not learn online. - Confirm through another person
Contact a relative, coworker, manager, or company representative. - Refuse unusual payment methods
Gift cards, cryptocurrency, and cash deliveries are major warning signs.
The FTC also warns in its family emergency scam guidance that criminals can create a cloned voice from a short audio sample found online.
How does this help someone become safer online?
Learning to verify instead of immediately trust helps someone become safer online because it works regardless of how realistic the AI content becomes.
This habit can protect:
- Banking and payment accounts
- Email and social media profiles
- Job-search information
- Family members
- Personal identity information
- Workplace systems
- Cryptocurrency wallets
- Passwords and authentication codes
The goal is not to decide whether every message was created by AI. The goal is to decide whether the request is legitimate before taking action.
What practical defenses should consumers set up?
Use several layers of protection:
- Create a family safe word: Use it during unexpected emergency calls.
- Enable multifactor authentication: Prefer passkeys, security keys, or authenticator apps.
- Use unique passwords: Store them in a reputable password manager.
- Limit public voice samples: Review who can see videos and livestreams.
- Make social profiles private: Remove birthdays, phone numbers, addresses, and family details.
- Verify payment changes: Confirm through an established phone number.
- Avoid unexpected links: Open official apps or type known addresses yourself.
- Freeze your credit when appropriate: This can limit new-account identity theft.
- Discuss AI scams with relatives: Make verification a normal family habit.
- Report suspected scams: Reporting helps investigators identify broader campaigns.
The FTC’s phishing guidance explains that scammers use believable stories to steal passwords, account numbers, and other personal or financial information.
What should you do if you fall for an AI scam?
Act quickly without blaming yourself. AI scams are designed to manipulate trust.
- Contact your bank or payment provider immediately
- Ask whether the transaction can be stopped or reversed
- Change exposed passwords from a trusted device
- Sign out of active account sessions
- Enable multifactor authentication
- Save messages, recordings, usernames, receipts, and phone numbers
- Notify the impersonated person or organization
- Report the incident to the FTC and FBI
- Monitor financial and online accounts
- Consider freezing your credit if identity information was exposed
AI makes scams more realistic, but it does not change the safest response. Slow down, leave the conversation, verify through a trusted channel, and never send money or sensitive information because a caller or video creates urgency.
