The first official count of AI-driven fraud
For 25 years, the FBI's Internet Crime Complaint Center (IC3) has published an annual report on online crime. The 2025 report, released in April 2026, is the first to include a section on artificial intelligence.
The headline numbers:
- 1,008,597 complaints in 2025, up from 859,532 in 2024
- Nearly $21 billion in reported losses
- 22,364 complaints tied to AI, with nearly $893 million in losses
- About $7.7 billion lost by Americans over 60, up 37% from 2024
The FBI describes how criminals are using AI: fake social media profiles, cloned voices, fabricated identification documents, and believable videos of public figures or loved ones.
Why $893 million is the floor, not the ceiling
The AI figure counts complaints where the victim or reporter identified AI as part of the scheme. Most people who get a convincing phone call from a "grandson" or a polished email from a "bank" have no way of knowing whether a machine wrote it or cloned the voice. So the real share of fraud that AI touched is almost certainly larger than what got labeled.
What we can see clearly is where AI fits. It makes impersonation cheaper, faster and more convincing, and impersonation is already the most reported fraud in the country.
Impersonation is the fraud Americans report most
The Federal Trade Commission's numbers for 2025 tell the same story from a different angle:
- 3 million fraud reports and $15.9 billion in reported losses, up from 2.6 million reports and more than $12 billion in 2024
- Impostor scams were the most reported fraud, as they have been since 2020: more than 1 million reports and more than $3.5 billion lost
- Investment scams caused the largest losses, at $7.9 billion
On the identity theft side, the FTC logged more than 1.3 million identity theft reports in 2025. The largest category was credit card fraud, covering hijacked accounts and new accounts opened in the victim's name, with 597,770 reports.
Put together: criminals pretend to be someone you trust, and AI is making that pretense harder to see through.
What AI-powered identity fraud looks like in real life
- The family emergency call. A short clip of someone's voice, from a voicemail greeting or a social video, can be enough to clone it. The call is urgent, emotional and asks for money fast.
- The fake "bank" or "fraud department." A caller or text that already knows your name, address and the last four digits of an account sounds legitimate. Those details often come from data broker listings or past breaches.
- New accounts in your name. With your Social Security number, date of birth and address, a criminal can apply for credit cards and loans. Fabricated ID documents make that easier.
- The convincing video. A deepfake of a celebrity pitching an investment, or of a relative asking for help.
Every one of these runs on personal data. The more of your information is for sale, the more convincing the attack.
Four things to do this week
- Agree on a family code word. If a call is urgent and emotional, ask for it. A cloned voice won't know it.
- Take a beat, then call back. The FBI's advice is to resist pressure to act quickly. Hang up and call back on a number you already have, not one the caller or message gave you.
- Freeze your credit. A free credit freeze at Equifax, Experian and TransUnion makes it much harder for anyone to open new credit in your name. You can lift it when you need to apply for credit.
- Shrink your public footprint. Data broker sites publish your address, phone numbers and relatives. That's the raw material for convincing impersonation. Removing it makes you a harder target. Our guide to removing your personal information from the internet shows how.
If it has already happened to you, follow our step-by-step plan for what to do if your identity is stolen, and report it at IdentityTheft.gov and IC3.gov.
Where Uncompromized fits
Uncompromized pairs the two defenses that matter most against AI-driven fraud: removing your information from 440+ data broker sites, and monitoring for your details showing up where they shouldn't. If fraud does happen, a US-based Recovery Concierge™ specialist handles the restoration, backed by $1M ID theft protection. For a deeper look at deepfakes, voice cloning and synthetic identities, read our guide to AI identity threats.