Elder Fraud + AI: same psychology, far more convincing deception

A recent Journal of Accountancy article highlights a trend many of us in cyber, risk, and fraud already feel daily: AI is making classic scams dramatically more believable and more scalable.

One case described an AI-enhanced “grandparent scam” where an 86-year-old victim lost $6,000 after believing she recognized her granddaughter’s voice on the phone.

That story perfectly illustrates the shift: the tactic isn’t new, the technology is.

Photo: Freepik

The article notes the “target market” is growing fast: people aged 65+ were 18% of the U.S. population in 2024 (up from 12.4% in 2004) and are projected to reach 20% by 2040.

At the same time, losses from elder fraud rose 43% to $4.89B in 2024, with phishing being the most frequently reported crime.

The article describes several AI-enabled techniques:

• Voice cloning (seconds of audio scraped from social media)

• Deepfake video/images impersonating family or officials

• AI-powered phishing that’s personalized and well-written

• Predictive targeting + chatbots that build trust over time (romance/investment scams)

It also cites an estimate that AI-generated fraud damages may reach $40B in the U.S. by 2027 (from $12.3B in 2023).

The best countermeasures are still human + process, supported by tools:

➡️ Verify out-of-band: hang up and call back using a trusted number

➡️ Use family “safe words” for emergencies

➡️ Enable scam call screening

➡️ Monitor accounts and set anomaly alerts (large/unusual transfers)

➡️ Tighten social media privacy settings to reduce audio/video exposure

None of this is truly “new.” Years ago, criminal groups exploited seniors with the classic “grandchild scam” over the phone. The technology changed, but the psychology didn’t: urgency, fear, isolation, trust in familiar voices.

What’s different now is the scale and sophistication: voice cloning, deepfakes, and highly personalized phishing will become more common and more convincing

Link

Author: Sebastian Burgemejster