- 75% of Americans believe seniors are more vulnerable to AI-powered scams than younger generations, yet they are the least prepared and least likely to report.
- The financial stakes are high for senior Americans with the reported average accessible savings and investments higher than any other age group at more than $103,000
One in three American seniors were targeted by AI scams in the last year, yet most households have not taken basic precautions to protect older relatives from emerging threats, according to a new survey from Sumsub, a global verification and fraud prevention platform.
The survey of 2,000 U.S. adults found that 75% of Americans believe seniors are more vulnerable to AI-powered scams than younger generations. One in three Americans (32.7%) said they personally know a senior family member or friend who was targeted by a financial scam in the past year, while 31.5% of seniors reported being personally targeted by a scam themselves.
The financial stakes are significant. Americans aged 55 and older reported average accessible savings and investments of more than $103,000, nearly three times the average held by respondents aged 18 to 24. Among seniors targeted by scams, one-third lost more than $2,000.
Yet the generation viewed as most vulnerable is also among the least prepared. Older Americans were the least likely to have discussed AI scams with family members, establish emergency verification plans such as family code words, or understand existing legal protections against AI-enabled fraud. They were also the least likely to report experiencing a scam attempt (28%), nearly twice as likely as the youngest adults to stay silent.
"What the data shows as the biggest risk for seniors is what happens after a scam takes place," said Danielle Labarbera, Vice President of Sales, North America at Sumsub. "They're less likely to talk to family, less likely to report it to authorities, and do not have any verification plan in place before an incident occurs. Awareness alone isn't closing that gap. The families who stay safe are the ones who discuss a protection plan before a fraudulent call or message takes place."
As AI-powered scams become more sophisticated, many Americans remain unsure if they could recognize them. Only 19.3% of Americans aged 55 and older believe they could identify an AI-generated cloned voice, compared to 60.8% of adults aged 18 to 24. Overall, just 34.5% of Americans say they are confident they could recognize an AI-generated voice clone.
Among respondents who encountered AI-enabled scams, the most common tactics included AI-generated emails or text messages impersonating trusted companies, fake customer service representatives, AI-generated bank fraud alerts, voice-cloning scams, and deepfake video messages.
As AI-powered scams become increasingly difficult to detect, Sumsub recommends families take proactive steps to protect themselves and elderly relatives. These steps include:
- Establishing a family emergency code word to verify urgent requests
- Treating unexpected requests for money or sensitive information with caution, even when the caller or person on screen appears to be someone you know
- Watching for potential signs of deepfakes, including unnatural facial movements or blinking, mismatched lighting and shadows, glitches around the hairline or jaw, audio that does not align with lip movements, and unusual pauses or changes in tone
- Ending suspicious interactions and independently contacting the person through a known phone number or trusted channel
- Enabling account and transaction alerts to quickly identify suspicious activity
- Reviewing social media privacy settings and limiting publicly available personal information that scammers can use to create convincing impersonations
- Reporting scams to financial institutions and authorities
- Having regular conversations with elderly relatives about emerging fraud tactics and what to do if they receive a suspicious request
For more information on how to spot deepfakes, visit Sumsub’s website.
The survey was conducted by Censuswide among 2,000 U.S. adults aged 18+ between June 15–19, 2026. Results were weighted to be nationally representative by age, gender, and region.
