Similar to how scammers are using artificial intelligence (AI) to impersonate CEOs to defraud businesses, they’re also mimicking regular people to prey on their elderly loved ones.
Case in point: an AI-enhanced grandparent scam that cost an 86-year-old grandmother in Philadelphia US$6,000 last year.
The woman received a phone call from a person she thought was her granddaughter. In the conversation that followed, the “granddaughter” said she had been detained by police following an accident, had been assigned an attorney, and needed money. The grandmother agreed to help her granddaughter out with $6,000 in cash, which the scammers picked up in person. They even gave her a receipt.
When the woman learned the accident was a ruse and her granddaughter was at work safe and sound, she said she would have never fallen for the scam had she not recognized her granddaughter’s voice on the phone.
Scammers have long preyed on the elderly to defraud them of money. The rise of generative AI has augmented these scams at a time when the number of potential victims is steadily rising.
The U.S. Census Bureau reported the percentage of the population age 65 and older increased to 18% in 2024, from 12.4% in 2004. By 2040, Americans age 65 and older are projected to account for 20% of the population.
The elderly population is also an attractive target. Americans age 65 and older:
- Have a greater accumulation of savings and retirement accounts, investment income, and other income streams.
- Are more likely to be struggling with increased isolation; anxiety and depression; an increased need for medical support; and cognitive decline and memory issues.
- Are increasingly uncomfortable with the general reliance on technology and are more likely to trust in what appear to be familiar voices, leaving them more vulnerable to voice cloning and deepfakes.
In 2024, losses from elder fraud rose 43% to $4.89 billion; phishing attacks were the most frequently reported crime. Deepfake phishing attacks are mostly delivered by email.
US $40 BILLION INDUSTRY TARGETING THE ELDERLY
The emergence of AI presents increased opportunities to create more realistic and nuanced messaging that not only raises the likelihood of success with these frauds but also increases the ability to disseminate the malicious messaging to a wider swath of the aging population.
AI-powered scams targeting the elderly use:
- Voice cloning: With just a few seconds of audio, often scraped from social media, scammers can clone the voice of a loved one (e.g., a grandchild) to make a convincing “family emergency” call. The caller typically claims to be in trouble (e.g., arrested, in an accident) and needs urgent, confidential financial help (e.g., immediate cash payment for bail), pressuring the victim to act before they can verify the story.
- Deepfake videos and images: AI can generate highly realistic fake videos or images of family members, public figures, or officials (e.g., Social Security or Medicare representatives) to gain trust and manipulate victims into revealing sensitive information or transferring money.
- AI-powered phishing: AI allows for the mass creation of personalized and grammatically correct emails or messages that appear legitimate, often impersonating banks, government agencies, or tech support. These messages can lure seniors into clicking malicious links, installing remote access software, or providing personal credentials.
- Predictive targeting and chatbots: Machine-learning algorithms can analyze data to identify the individuals most susceptible to fraud based on their online behavior and demographics. AI-powered chatbots can then engage in realistic, long-term conversations, as seen in romance or investment scams (also known as pig butchering scams), to build trust and eventually trick victims into sending large sums of money.
Deloitte’s Center for Financial Services estimated that AI-generated fraud will reach $40 billion in damages in the United States by 2027, a 32% annual increase from $12.3 billion in 2023.
HOW TO DETECT AI-POWERED FRAUD
The ways to detect and prevent possible fraud against the elderly don’t differ much whether the scams are AI-powered or more traditional. They include:
- Bank and credit monitoring: A trusted family member or adviser should frequently review bank and credit card statements to look for large and/or unusual transactions. Signing up for a credit monitoring service can assist in blocking unauthorized attempts for new credit or spot instances where new credit was obtained.
- Alerts to changes in behavior: A trusted adviser can establish what is a “normal” withdrawal or transfer of money for an individual. This will help raise red flags in a review of behaviors.
- Scam/spam call screening: Telephone providers can give users access to tools that help detect spam calls to cellphones and landlines and can send them directly to voicemail or block them. People should also be encouraged to answer calls from only known numbers/callers.
- Use of “safe” words: Family members can use a safe word to identify themselves and thereby confirm that they can be trusted.
