The FBI’s Internet Crime Complaint Center gave artificial intelligence its own fraud category for the first time in the bureau’s nearly 25-year history of tracking online crime, and the toll on older Americans is already substantial. In its newly published 2025 Internet Crime Report, IC3 recorded more than 3,100 complaints from people age 60 and older that referenced AI, with reported losses topping $352 million. Those cases involve cloned voices, fabricated videos and AI-written messages built to make a stranger, or a familiar-sounding relative, seem real enough to trust.
A New Category, and a Number Specific to Older Victims
Nationally, IC3 logged 22,364 complaints in 2025 carrying an “AI Related” descriptor, with adjusted losses exceeding $893 million, according to the bureau’s 2025 Internet Crime Report. Of those, 3,143 complaints came specifically from people 60 and older, with $352,496,231 in reported losses attached to that age group — the exact figures behind the numbers now circulating about the report.
The FBI’s own announcement of the findings singled out the same trend. Cryptocurrency- and artificial-intelligence-related complaints were “among the costliest” categories in a year that saw overall cyber-enabled fraud losses approach $21 billion, the bureau said in an April 2026 press release tied to the report’s release. IC3 receives close to 3,000 complaints a day, and 2025 was the first year the agency broke AI out as its own tracked descriptor rather than folding it into whatever underlying scam type — romance, investment, tech support — it happened to support.
Free retirement updates: Miss an enrollment or claim deadline and it may be gone. Our free Retirement Shield newsletter keeps readers ahead of the ones that matter. Get the free newsletter.
How Scammers Use Cloned Voices and Fake Videos
The report walks through specific tactics rather than treating AI as one generic threat. Chat generators can produce official-sounding emails that mimic a company executive, sometimes carrying phishing links or wiring instructions — a twist on the business email compromise schemes IC3 has tracked for years. Voice-cloning tools let a scammer imitate a real person’s voice closely enough to request an urgent wire payment over the phone. Deepfake video and audio can be built to impersonate a family member, often deployed inside a fabricated emergency to pressure someone into sending money before they have time to check the story with anyone else.
None of this requires movie-studio resources. A short clip pulled from a voicemail greeting or a public social media video can supply enough raw audio to build a usable voice clone. That is a large part of why the bureau treats AI less as a brand-new crime and more as a force multiplier bolted onto scam types it already tracks, from grandparent scams to romance schemes to tech-support fraud.
The Elder Fraud Backdrop These Cases Sit Inside
The AI numbers sit inside a much larger elder-fraud picture. IC3 counted 201,266 complainants age 60 and older in 2025, reporting roughly $7.7 billion in losses across every fraud type combined — up 37% from 2024, according to the FBI. More than 12,000 of those complainants each lost more than $100,000 in a single year. Against that backdrop, the AI-specific numbers are a small slice of the total complaint count but a fast-growing one, since the descriptor did not formally exist as its own tracked category in prior years’ reports.
Why AI Cases Run Nearly Three Times the Elder-Fraud Average
Divide the dollar figures and a pattern emerges. Nationally, the average AI-flagged complaint involved a loss of roughly $40,000. Among the 3,143 complaints from people 60 and older, the average ran closer to $112,000 — nearly three times higher, and well above even the $38,500 average loss the report records across every type of elder-fraud complaint combined. That gap suggests AI tools are showing up disproportionately inside the highest-dollar scam categories, like investment fraud, rather than in smaller-dollar cons.
A Campaign Built Before AI Had Its Own Category
The FBI’s “Take a Beat” messaging predates this year’s AI numbers. The bureau launched the nationwide campaign in August 2024, citing $1.6 billion in fraud losses in just the first five months of that year, and built its core advice around the same pressure tactics — urgency, isolation, promises of money or companionship — that AI tools now make easier to fake convincingly. The campaign’s practical steps still apply directly to AI-enabled scams: verify unknown contacts independently, search for a business or caller’s name before responding, and never send money, gift cards or wire information to someone you have not independently verified.
What the FBI Is Telling Older Adults and Their Families to Do
The bureau’s public message is deliberately simple: “Take a Beat.” Resist pressure to act immediately, verify who is actually on the other end of a call or video, and confirm any urgent money request through a phone number you already know, never one supplied by the caller or the email itself. The FBI’s common frauds and scams guidance also states plainly that the bureau will never call or email a private citizen asking them to move money by wire, cryptocurrency, gift card or prepaid card. Families with older relatives can use the same rule: agree in advance on a phrase or callback habit for any request involving money, so a cloned voice on the phone has less chance of working.
Filing a report matters even after the money is gone. IC3 says rapid reporting can support fund-recovery efforts in some cases and helps investigators connect individual complaints to larger criminal patterns — patterns that, for the first time in 2025, included artificial intelligence as a tracked variable in its own right. The bureau’s report frames this as an early snapshot rather than a peak: AI tools are getting cheaper and easier to use, and 2025 is the baseline year the FBI will now measure every future report against.
This article was produced with AI assistance and reviewed by a human editor. Figures are linked to their primary sources; where a claim could not be verified from the public record, we say so.
More Financial Reading
