Reader-supported: we may earn a commission from links on this page.
A phone call in your grandson’s voice. A video call with your boss. A perfectly written email from your bank. Scammers now use AI to fake all three, and it’s working. In its 2025 Internet Crime Report, the FBI tracked AI-related complaints for the first time and counted 22,364 of them, with losses of nearly $893 million.
We went through the latest warnings from the FBI, FTC, FCC and CFTC and pulled out the 6 AI scams to know, the red flags they share, and the simple habits that stop them.
We research, we don’t lab-test. Everything below comes from the official reports, warnings and news reports listed at the end, checked in October 2026.
In this guide:
- AI Scams: The Short Answer
- How Big Is the Problem?
- The 6 AI Scams to Know
- Red Flags: How to Spot an AI Scam
- How to Protect Yourself From AI Scams
- What to Do If You’ve Been Scammed
- Our Verdict
- Frequently Asked Questions
AI Scams: The Short Answer
AI doesn’t invent new scams. It makes old ones convincing. The stories are the same as ever: a loved one in trouble, a boss who needs a payment today, an investment that can’t lose. What AI changes is the voice, the face and the writing, which can now look and sound real. The defense hasn’t changed: slow down, hang up, and check through a phone number or website you already trust. A family safe word and a firm rule against paying by gift card, crypto or wire stop most of these scams.
How Big Is the Problem?
AI scams are still a small slice of fraud overall, but they sit on top of a fraud problem that keeps growing:
- Nearly $893 million: what Americans reported losing to AI-related scams in 2025, across 22,364 complaints, according to the FBI. It’s the first year the FBI’s annual report has had an AI section.
- Nearly $21 billion: total cybercrime losses reported to the FBI in 2025, from 1,008,597 complaints.
- $7.7 billion: losses reported by Americans over 60, up 59% from 2024. Their complaints rose 37%.
- $15.9 billion: fraud losses reported to the FTC in 2025, from 3 million reports. Investment scams caused the biggest losses ($7.9 billion), and imposter scams were the most reported, with more than 1 million reports.
These are only the losses people report. Many victims never file a report, so the real numbers are higher.
The 6 AI Scams to Know
Each scam below comes from an official warning or a documented case.
1. Voice-clone “family emergency” calls
Your phone rings and it’s your daughter, grandson or partner, panicking. They’ve been in a car accident, arrested or kidnapped, and they need money right now. Please don’t tell anyone.
It’s an old scam with a new trick. The FTC warns that a scammer only needs a short audio clip of your family member’s voice, which they can pull from videos posted online, plus a voice-cloning program. The FBI says criminals use these clips to pose as a relative in a crisis and demand money or a ransom.
How to beat it: Hang up and call your family member on the number you already have for them. If you can’t reach them, call someone close to them. Better yet, agree on a family safe word now (more on that below). Also know that robocalls using an AI-generated voice are illegal in the US unless you agreed to them in advance, under a 2024 FCC ruling.
2. Deepfake bosses and officials
In early 2024, a finance worker at the engineering firm Arup in Hong Kong joined a video call with people who looked and sounded like the company’s chief financial officer and other colleagues. All of them were deepfakes. The worker sent 200 million Hong Kong dollars, about $25.6 million, across 15 transactions, CNN reported.
The same playbook targets individuals. The FBI warned in December 2025 that criminals have been impersonating senior US officials since at least 2023, using texts and AI-generated voice messages. They chat briefly, then ask to move to Signal, Telegram or WhatsApp. There, they ask for a verification code, ID documents, money or an introduction to someone you know. The FBI notes that AI voices can sound nearly identical to a person you know.
How to beat it: Treat any surprise request for money, codes or documents as unverified until you confirm it through a phone number or email you found yourself. Never share a two-factor code with anyone. At work, require a call-back on a known number before any urgent or secret payment.
3. AI-written phishing texts and emails
Bad spelling used to give scammers away. Not anymore. The FBI says criminals use AI to write messages faster and to translate them, which removes the grammar and spelling mistakes that once exposed scammers overseas. Some fake websites even use AI chatbots to steer visitors toward malicious links.
How to beat it: Judge a message by what it asks you to do, not by how well it’s written. Don’t click links in unexpected texts or emails; open the website or app yourself. And protect your most important accounts, especially email, with a passkey or a hardware security key. Both only work on the real website, so a fake login page can’t steal them. We compare the options in our guide to which security key you need.
4. Deepfake investment pitches and “AI trading bots”
Investment fraud is where the biggest money goes: the FBI says it accounts for nearly 49% of all scam-related losses. AI adds fake promotional videos, celebrities who seem to endorse a platform, and polished websites for fake crypto investments, all uses the FBI lists.
Many pitches promise an “AI trading bot” that makes money for you. The CFTC, which regulates US futures markets, warns that AI can’t predict the future or sudden market changes. Its example: Mirror Trading International took more than $1.7 billion in bitcoin from at least 23,000 people with a bot that promised at least 10% a month.
How to beat it: Never invest through an ad, a celebrity video or someone you met online. Guaranteed or very high returns are a red flag every time. Before you send money, look up the firm and the person on FINRA BrokerCheck.
5. Fake AI apps and tools that install malware
Scammers also use the AI craze itself as bait. In May 2025, Google’s Mandiant team reported a campaign that ran thousands of Facebook ads, and about 10 on LinkedIn, for fake “AI video generator” websites. More than 30 sites copied real tools such as Luma AI, Canva Dream Lab and Kling AI. Whatever you typed in, the site handed you malware, sometimes disguised as a video file, that could steal passwords, browser cookies, card details and crypto wallets.
How to beat it: Get AI tools only from the developer’s official website or your phone’s app store, never from an ad. Be suspicious of any “video” that arrives as a .zip or .exe file. And before you connect any AI tool to your email or bank, check what it can access. We did that for Meta’s AI agent, Muse.
6. Fake “fund recovery” help, including deepfakes of FBI leaders
This one targets people who have already lost money. In a July 2026 warning, the FBI said scammers pose as FBI or IC3 staff on Facebook Messenger and Telegram and offer to help recover lost funds. Some use AI-generated videos of a senior FBI leader urging people to file complaints on a fake IC3 website. That site collects your name, phone number, email, scam type and how much you lost.
How to beat it: The IC3 has no social media accounts, never contacts people directly by phone, email or social media, and never charges to recover money. Type ic3.gov into your browser yourself, and skip sponsored search results.
Plus: AI romance scams and catfishing
Scammers also use AI-made photos, chatbots and voice clones on dating apps and social media. We cover those in our guide to AI romance scams, including how to reverse image search a suspected catfish for free. To lock down your own dating profile, see Bumble vs Hinge privacy settings.
Red Flags: How to Spot an AI Scam
Whatever the AI trick, the scam still needs you to act fast and pay in a way that’s hard to undo. Watch for these signs:
| Red flag | What it looks like |
|---|---|
| Pressure to act now | “Send it in the next hour or I’ll be arrested.” Strong fear or excitement is the point. |
| Secrecy | “Don’t tell Mom.” “This deal is confidential.” |
| Hard-to-reverse payment | Gift cards, crypto, wire transfers or a crypto ATM. |
| Switching apps | A request to move the chat to Signal, Telegram or WhatsApp. |
| Asking for codes | Any request for a verification or two-factor code. |
| Guaranteed returns | “AI trading bot,” “risk-free,” or returns of 10% a month. |
| Audio or video glitches | Odd hands or teeth, mismatched shadows, voice lag, strange tone or word choice. |
A note on glitches: the FBI lists them, but don’t count on spotting them. A cloned voice can sound just like the real person, and the fakes keep getting better. Checking the story through a channel you trust works every time. Spotting a deepfake doesn’t.
How to Protect Yourself From AI Scams
- Set a family safe word. The FBI recommends a secret word or phrase your family can use to prove who’s calling. Pick something that isn’t on social media, share it in person, and never send it by text.
- Hang up and call back. If a call, text or video chat asks for money or information, end it and contact the person or company through a number you already have.
- Never pay strangers with gift cards, crypto or wire transfers. No real government agency, bank or relative in trouble needs you to pay this way.
- Make your voice and face harder to copy. Set your social media accounts to private, only accept followers you actually know, and think twice before posting long videos of yourself talking.
- Lock down your accounts. Turn on two-factor authentication and never share the codes. For your email, use a passkey or a hardware security key such as the Yubico Security Key C NFC. Our security key guide explains which one fits your phone and laptop, and why you should buy two.
- Shrink your data footprint. Scammers personalize their stories with details from data broker sites, like your relatives’ names, your address and your phone number. Our data broker removal guide shows how to opt out for free.
- Freeze your credit. It’s free, and it stops scammers from opening new accounts in your name. See USA.gov’s credit freeze guide.
Want extra help with scam texts and phishing? Surfshark’s One+ plan includes scam protection that flags suspicious texts, checks your Gmail for phishing and blocks known phishing sites. It also includes breach alerts and Incogni, which removes your details from data broker sites (for US, UK, EU, Swiss and Canadian residents). It doesn’t screen phone calls, so the safe word and call-back habits above still matter. Check the current price and which features work on your devices before you buy.
What to Do If You’ve Been Scammed
- Stop paying and cut contact. Scammers often come back for a second payment.
- Call your bank or payment app right away. Ask them to stop or reverse the payment. The sooner you call, the better your chances.
- Report it. File a complaint with the FBI at ic3.gov and with the FTC at ReportFraud.ftc.gov. If you’re 60 or older, the Justice Department’s Elder Justice Hotline can help: 1-833-FRAUD-11 (833-372-8311).
- Write everything down. The FBI asks for the scammer’s name or company, how they contacted you, the dates, how you paid, and where the money went.
- Secure your accounts. If you shared a password or code, change the password and turn on two-factor authentication. If you shared your Social Security number, freeze your credit.
- Watch for recovery scams. Anyone who contacts you offering to get your money back for a fee is almost certainly running scam number 6.
Our Verdict
AI has made scams look and sound real, but it hasn’t changed what they need from you: speed, secrecy and a payment you can’t take back. That’s good news, because the best defenses are simple and free. Set a family safe word this week, make “hang up and call back” a habit, and never pay a stranger in gift cards, crypto or wire transfers. Those three habits beat a cloned voice just as well as an old-fashioned phone scam.
Want the deeper investigation? AI scams are one of the topics in The Online Predator Bible, our full investigation into AI scams, data brokers and dating app privacy. See all our privacy ebooks.
Frequently Asked Questions
What are AI scams?
AI scams are frauds that use artificial intelligence to make a lie more convincing: cloned voices, deepfake videos, AI-made photos and IDs, or flawless phishing messages. The goal is the same as any scam: to get you to send money or share personal information.
How do AI voice cloning scams work?
A scammer takes a short clip of someone’s voice, often from a video posted online, and uses a voice-cloning program to make it say anything. Then they call a relative, pretend to be in trouble and ask for money right away. Hang up and call the person back on a number you know.
How can you tell if a call is an AI voice?
Listen for an odd tone, strange word choices or a lag before answers, which the FBI lists as warning signs. But clones can sound just like the real person, so don’t rely on your ears. Ask for your family safe word, or hang up and call back.
What is a family safe word?
It’s a secret word or phrase that only your family knows, used to confirm who is really calling. The FBI recommends one. Choose something that isn’t on social media and share it in person.
Are AI robocalls illegal?
In most cases, yes. In February 2024 the FCC ruled that AI-generated voices count as artificial voices under the Telephone Consumer Protection Act, so robocalls that use them need your consent in advance.
How much money do people lose to AI scams?
In 2025 the FBI received 22,364 complaints related to AI, with reported losses of nearly $893 million. That only counts what people reported, so the real figure is likely higher.
Where do I report an AI scam?
Report it to the FBI’s Internet Crime Complaint Center at ic3.gov and to the FTC at ReportFraud.ftc.gov. If you sent money, call your bank or payment app right away.
Sources
- FBI: Cryptocurrency and AI Scams Bilk Americans of Billions (April 6, 2026)
- FBI IC3: 2025 Internet Crime Report
- FBI IC3: Criminals Use Generative Artificial Intelligence to Facilitate Financial Fraud (December 3, 2024)
- FBI IC3: Senior U.S. Officials Continue to be Impersonated in Malicious Messaging Campaign (December 19, 2025)
- FBI IC3: FBI Warns of Scammers Impersonating the IC3 (July 20, 2026)
- FTC: Scammers use AI to enhance their family emergency schemes (March 20, 2023)
- FTC: Testimony to the Joint Economic Committee on 2025 fraud losses (March 25, 2026)
- FCC: FCC Makes AI-Generated Voices in Robocalls Illegal (February 2024)
- CFTC: AI Won’t Turn Trading Bots into Money Machines
- CNN: Arup revealed as victim of $25 million deepfake scam (May 2024)
- Google Cloud / Mandiant: Text-to-Malware, fake AI video generator websites (May 2025)
- USA.gov: How to freeze your credit
