A few years ago, a scammer needed time, effort, and some level of skill to trick someone out of money.
Today, they need software. With widely available AI tools, an attacker can clone a voice, generate a convincing video, and build a believable identity in minutes. What used to require coordination and expertise can now be done at scale, often by a single individual.
That shift is changing who gets targeted most. Older adults are increasingly at the center of this new wave of fraud, and the scale of the problem is growing faster than most people realize.
A crisis measured in tens of billions
The numbers are staggering. Americans aged 60 and older lost an estimated $81 billion to fraud in a single year. And that figure likely understates the true impact. Many cases go unreported, often because victims feel embarrassed or unsure where to turn.
At the same time, the nature of these attacks is evolving quickly. AI-enabled fraud has increased by over 400% in just two years (2023-2025), according to recent industry analysis, driven by rapid improvements in voice cloning, deepfake video, and AI-generated impersonation.
One phone call can be enough
To understand why this threat is so dangerous, it helps to look at what one of these attacks feels like in practice.
Two years ago, Gary Schildhorn received a frantic phone call that sounded exactly like his son. The voice said he had been in an accident and needed money immediately to post bail. Gary was on the verge of wiring thousands of dollars when he stopped and called his son directly. His son answered, safe at home and completely unaware of the call.
The original voice had been cloned using AI. Gary avoided becoming a victim by pausing for a moment and verifying the story through another channel. But the story is powerful because it shows how believable these attacks have become. This was not a sloppy scam filled with obvious warning signs. It sounded personal. It sounded urgent. It sounded real. That is exactly why these attacks work.
Why older adults are being singled out
There is a tendency to assume scams succeed because someone made a careless mistake.
That view misses the point. Modern scams are designed to override normal decision-making. They rely on urgency, emotional pressure, and familiarity. Older adults are often more likely to answer unknown calls and engage in conversation. Many have built savings over decades, which makes them attractive targets. And many have not been trained to expect AI-generated deception that can mimic a loved one’s voice or a trusted authority figure.
Attackers understand all of this. Instead of sending generic messages, they create situations that feel deeply personal. A grandchild in trouble. A bank calling about suspicious activity. A police officer asking for immediate action. A new romantic partner building trust over weeks or months. That sense of normalcy is what makes the fraud effective.
From obvious scams to believable stories
Traditional scams often gave themselves away. The grammar was poor. The request was strange. The email address looked suspicious.
AI is removing many of those clues. Today’s attackers can clone a voice from a short audio sample, generate highly persuasive messages, and impersonate family members, financial institutions, or government officials with alarming accuracy. The attack works because the story feels credible. It feels personal, urgent, and familiar.
That is an important shift. We are seeing attacks that involve:
- a family member supposedly calling in distress
- a government official demanding immediate payment
- a romance scam that unfolds over months
- coordinated impersonation across phone, text, and video
These scams are effective even against people who are informed, because the interaction is carefully staged to feel authentic from the very first moment.
The cost goes beyond money
Financial loss is only one part of the harm. Victims often experience shame, anxiety, and a lasting loss of confidence in their own judgment. Some begin to distrust ordinary forms of communication. Others withdraw socially after realizing how convincingly they were manipulated.
These incidents also affect families in lasting ways. They can strain relationships and leave loved ones to manage the confusion, emotional stress, and leave loved ones to manage the financial and administrative fallout.
Why education matters most right now
There is no single tool that will stop every AI scam. The technology will continue to improve. Detection systems will reduce risk, though some attacks will still get through. At the same time, regulation is unlikely to keep pace as these tactics evolve rapidly across borders.
The most practical defense available today is awareness. That is one reason we developed Protecting Older Adults from AI Scams, a public education course designed to help families recognize the warning signs of AI-enabled fraud and respond before money is lost. The goal is simple: give older adults and their loved ones practical tools they can use in real situations.
People need simple habits they can rely on in high-pressure moments. In our experience, a few actions consistently reduce risk:
- pause when a request feels urgent
- verify the story through a separate, trusted channel
- never send money under pressure
- create family verification steps, such as a shared code word
- avoid making sensitive decisions alone during an emotional moment
While these habits are simple, they matter because they create friction. And friction is often enough to break the scam.
Gary Schildhorn’s story is a perfect example. What saved him was not advanced technology. It was one decision to stop and make a second call. The same is true outside the workplace. Awareness only works when it feels relevant to the person receiving it.
A broader shift in how we think about security
For years, cybersecurity has focused on protecting systems and organizations. Although that focus still matters, today’s risks also require protecting individuals in their personal lives.
Some of the most damaging attacks now target people directly. Older adults are experiencing this shift early and at high volume, with broader exposure likely to follow.
As AI tools become cheaper and easier to use, these attacks will spread further. As Stanford behavior scientist BJ Fogg has shown, people are more likely to act when a message combines motivation, ability, and a prompt in the same moment. In scams, that can mean a familiar voice, a simple request, and a sense of urgency arriving all at once.
What families should do now
The first step is to talk about this openly. Families should discuss how these scams work before a crisis happens. Parents and grandparents should know that a familiar voice on the phone is no longer proof of identity. Adult children should encourage simple verification habits and make it easy to check a suspicious request without embarrassment.
Preparedness and support should be the goal of those conversations. AI-driven fraud is becoming part of everyday life. The sooner families recognize that, the better positioned they will be to stop it.
About the Author
Brian Long is the CEO and Co-Founder of Adaptive Security, where we build software to protect companies from new cyber threats like deepfakes, GenAI phishing, smishing and more. Today, Adaptive is growing fast, working with leading businesses like regional and national banks, leading tech and software companies, and critical health systems.
Brian can be reached online at https://www.linkedin.com/in/brianclong/ and at our company website https://www.adaptivesecurity.com/
