Scammers armed with artificial intelligence can now clone a grandchild’s voice from a short social media clip and place a panicked phone call demanding cash, and the FBI says this tactic is driving one of the fastest-rising fraud categories of the summer. The bureau’s 2025 Internet Crime Report ties voice-cloning software to billions of dollars in AI-enabled losses, while older adults remain the primary targets, accounting for more than 201,000 complaints and $7.7 billion in total fraud losses in the most recent reporting cycle. The shift from a crude impersonation script to a synthetic voice that sounds nearly identical to a loved one is compressing the time victims have to think before they pay.
Why cloned-voice grandparent scams are accelerating now
The traditional grandparent scam follows a familiar pattern: a caller pretends to be a grandchild in legal trouble, begs for secrecy, and asks for wire transfers or mailed cash. Between January 2020 and June 2021, the FBI’s Internet Crime Complaint Center logged 650 reports and more than $13 million in losses from this scheme, with 90 victims losing more than $3.6 million through at-home cash pickups alone. A later IC3 window, January through September 2023, recorded 195 complaints and at least $1.9 million in losses from the same category. Those numbers, already significant, reflect a period before voice-cloning tools became cheap and widely accessible.
What changed is the barrier to entry. The Federal Trade Commission warned that scammers can use a short audio clip pulled from online content and pair it with voice-cloning software to sound like a loved one. A grandchild’s TikTok video or Instagram story can supply enough raw audio for the software to generate a convincing fake in seconds. The FBI stated in a 2025 cyber alert that “malicious actors are more frequently exploiting AI-generated audio to impersonate personal relations” and that AI voice cloning “can sound nearly identical.” That realism collapses the natural skepticism that once gave victims a few seconds to question whether the caller was really family.
Summer amplifies the risk for a practical reason: school breaks and travel plans put grandchildren out of easy reach, making it harder for an older adult to quickly verify whether a grandchild is actually in trouble. Scammers exploit that gap, pressuring victims to act before they can call back or check with other family members.
FBI and FTC data behind the voice-cloning threat
The FBI’s 2025 Internet Crime Report serves as the primary dataset linking cryptocurrency and AI scam losses to billions of dollars stolen from Americans. Within that broader picture, older adults bear a disproportionate share: victims age 60 and older filed more than 201,000 complaints, reported losses exceeding $7.7 billion, faced an average loss above $38,000 per incident, and included 12,400 individuals who each lost more than $100,000. Those figures span all fraud types, but the FBI has singled out AI-generated audio as a growing driver of the impersonation schemes that hit seniors hardest.
IC3 separately detailed how generative AI increases the believability and speed of fraud operations. The technology enables faster message generation, translation that strips away the language errors victims once used as red flags, and the creation of synthetic profiles that make scammers harder to identify. During RSA Conference week in May 2024, the FBI’s San Francisco field office issued a public warning about voice and video cloning scams as part of a broader alert on AI-enabled social engineering. On the regulatory side, the FTC finalized its Government and Business Impersonation Rule and proposed new protections specifically targeting AI impersonation of individuals, signaling that enforcement agencies view synthetic voice fraud as a distinct and growing enforcement priority.
The combined federal response, spanning IC3 complaint tracking, FBI field-office warnings, and FTC rulemaking, reflects an institutional consensus that cloned audio has changed the economics of impersonation fraud. A scam that once required a convincing actor now requires only a laptop and a few seconds of stolen audio.
Gaps in the data and what families should do first
A significant blind spot remains in the public record. Neither the FBI nor IC3 has published a quarterly breakdown that isolates complaints involving confirmed voice cloning from those using traditional audio impersonation. Without that separation, it is not yet possible to measure exactly how much higher the average loss climbs when a victim hears a cloned voice versus a live impersonator. The hypothesis that AI-voice cases produce larger per-victim losses, driven by higher compliance rates, is consistent with the FBI’s own language about added realism, but the data to prove it has not been released in a form that allows independent analysis.
The DOJ Elder Justice Initiative hotline statistics, similarly, are not broken out by AI versus non-AI variants. That means the full scale of cloned-voice fraud against older adults is almost certainly undercounted in current public reporting.
More from Morning Overview
*This article was researched with the help of AI, with human editors creating the final content.
