AI Voice Scams: How Fraudsters Target Families

Fraudsters clone a family member's voice from short clips scraped off social media, then use that voice to convince elderly relatives to send money during a fabricated emergency. The scam works because the voice sounds real. Under Section 415 of Israel's Penal Law, this is obtaining property by deception, and courts treat crimes against elderly victims more severely.

How Does the AI Family Voice Scam Work?

In my Tel Aviv criminal defense practice, I have seen a growing number of cases in recent months where fraudsters use artificial intelligence to clone the voice of a family member. The cloned voice is used to persuade an elderly relative to transfer money urgently, a digital escalation of the long-running "grandchild in distress" scam.

The typical call: an elderly person hears "their son" or "their grandson," sounding distressed, claiming a car accident, an arrest, or urgent financial trouble, and asking for money right away through a payment app, a prepaid card, or cryptocurrency. The emotional pressure and the sense of urgency are the tools of the manipulation. They are designed to deny the victim time to think or to verify the story with the actual family member.

The scale of the problem is growing everywhere, not only in Israel. The FBI's Internet Crime Complaint Center linked AI-related fraud to 22,364 complaints and close to $893 million in reported losses in its 2025 report, the first time in the center's history that artificial intelligence warranted its own category. Israeli police do not yet publish a comparable breakdown, but the mechanism reported to my office matches the pattern described in that report almost exactly.

How Does the Voice Cloning Technology Actually Work?

Modern voice cloning tools need only a few seconds of a person's recorded speech, often pulled from a video posted to Instagram, TikTok, or a wedding clip shared on Facebook. The software maps the pitch, cadence, and accent of that sample, then generates new sentences the real person never said. The victim is not hearing an old recording. They are hearing a synthetic voice invented for the call, which is why the scam survives even when the real family member is asleep, at work, or unreachable at the moment the call comes in.

This is the same underlying technology behind a widely reported 2024 case in Hong Kong, where police said an employee at the engineering firm Arup was tricked into wiring $25.6 million after joining a video call in which every other participant, including the company's chief financial officer, was an AI-generated deepfake. That case involved video and voice together and targeted a company rather than a family, but it shows how far the same cloning technique now reaches: from a single elderly victim on a phone call to a multi-million-dollar corporate wire fraud.

What Israeli Law Applies to AI Voice Cloning Fraud?

Legally, this is obtaining property by deception under Section 415 of Israel's Penal Law. Depending on how the call is built, prosecutors may add charges of forgery, impersonation, and unlawful use of a database. Where digital means carry out the act, offenses under the Computers Law can apply as well. Penalties rise when the victim is elderly, because Israeli law recognizes the particular vulnerability of that population.

Because the fraud usually involves a cross-border payment app, a prepaid card issued abroad, or cryptocurrency, these files often carry a money laundering dimension in addition to the fraud itself. That combination changes both the investigative path and the exposure for anyone who unknowingly moves the funds onward, a pattern I have written about in the context of ordinary Israeli businesspeople who get pulled into laundering files without criminal intent.

How Can Families Protect Themselves in Advance?

My standing recommendation to families is to agree on a code word in advance, one that never appears on social media and is unknown to strangers, and to use it to verify identity during any urgent call. Parents and grandparents should build the habit of hanging up and calling the family member back on a known number, rather than continuing the suspicious call itself.

Agree on a private family code word that is never posted online or said aloud on a recorded call.
Hang up on any urgent money request and call the family member back on a number you already have saved.
Never send money, gift cards, or cryptocurrency while still on the call that requested it.
Treat a refusal to wait a few minutes for verification as the clearest sign of fraud, not as proof of urgency.
Limit public video and voice clips of elderly relatives and children where possible, since a few seconds of clear audio is enough to clone a voice.

No financial decision should be made under pressure. A legitimate emergency will tolerate a few minutes of verification. I wrote about similar pressure tactics in a previous guide to cyber fraud, and the same rule applies here: urgency is the tell, not the emergency.

What Should You Do If You Have Already Been Targeted?

Anyone who has transferred money, or suspects an attempted fraud, needs to act fast. Contact the bank or the transfer company immediately and ask for the transfer to be stopped, since in some cases the funds can still be halted before withdrawal. File a complaint with the police, either at the nearest station or with Israel Police's National Fraud Investigations Unit, and preserve every piece of evidence: a recording of the call if one exists, the calling number, transfer confirmations, and any correspondence.

A parallel complaint can also be filed with Israel Police's national cyber unit, which specializes in offenses of this kind. Legal representation at the complaint stage helps ensure the complaint is drafted to maximize the chance of an investigation, and in appropriate cases, a parallel civil action for recovery of the funds can be considered. Where the transfer moved through a platform's payment tools, there can also be a separate liability question for the platform itself, a theory now being tested in Israeli courts in a pending lawsuit against Meta over fraudulent advertisements. Where the funds moved into a crypto wallet before the victim or the bank could act, recovery becomes harder still, for the same reasons I outlined in a guide to appealing frozen crypto assets.

The Warning Sign Families Should Never Ignore

The technology behind these scams will keep advancing, but the response that works has not changed: an open conversation with parents and grandparents, a family code word, and the habit of hanging up to verify. Extreme urgency, more than the voice itself, is the clearest sign of fraud.

More from Sagiv Rotenberg

Fraud protection

Also from Sagiv Rotenberg

All essays: Sagiv Rotenberg on Olam.

Sagiv Rotenberg (עו"ד שגיב רוטנברג) is a Tel Aviv criminal defense attorney with more than 25 years of practice. He is a founding partner of Rotenberg Criminal Law Office, which he leads with his brother, Nir Rotenberg. His practice focuses on fraud, financial crime, and cross-border criminal exposure facing internationally active Israeli business figures.