PL
HU
EN
Just five years ago, creating a realistic fake video required a team of specialists, professional software, and weeks of work. Today, all it takes is a laptop, a free app, and about ten minutes. Artificial intelligence has democratised the creation of fake content | and unfortunately, scammers have embraced this technology with open arms.
In this article, we'll show how AI is changing the landscape of scams targeting older adults, what new threats have emerged over the past two years, and | most importantly | how to defend against them.
Voice cloning
Voice cloning is a technology that not long ago sounded like science fiction. Today it's frighteningly accessible. Modern AI tools need just three to five seconds of someone's speech to create a synthetic copy of their voice. This copy can say any sentence | with an intonation, accent, and emotion resembling the original.
Where do scammers get voice recordings? Sources are everywhere. A video posted on Facebook, a voice message on WhatsApp forwarded on by someone, a recording from a family gathering shared on YouTube, or even a short statement in local TV footage. A few seconds is enough | the algorithm does the rest.
In Hungary, the news site G7.hu reported a series of 2024 cases in which scammers used cloned voices to defraud older people of money. The pattern is usually the same: the phone rings, and on the other end is the voice of a son or grandson saying he's had an accident, is in the hospital, and urgently needs tens of thousands of forints. The victim hears a familiar voice | and reacts emotionally, not rationally. That's exactly what scammers are counting on.
What's especially dangerous is that voice cloning is practically undetectable by the human ear. Even people who know a loved one's voice well can't tell the original from a synthetic copy. In laboratory studies, listeners' accuracy at spotting a fake voice is little more than 50% | the same as random guessing.
Video deepfakes
If voice cloning is frightening, video deepfakes take the threat to a whole new level. Deepfake technology lets you superimpose one person's face onto another's body in a video recording | in a way that's practically undetectable to the naked eye.
Scammers increasingly use deepfakes to create fake ads featuring celebrities. In Poland, deepfake videos have appeared in which well-known figures | actors, TV presenters, politicians | supposedly recommend investment platforms, miracle cures, or money-making schemes. The recordings look professional, the lips move in sync with the audio, the expressions look natural. For someone who doesn't know what a deepfake is, this kind of material is indistinguishable from the real thing.
The problem is especially serious among older adults. CSIRT KNF research from 2024 shows that people over 60 have significantly more difficulty recognising deepfake material than younger age groups. This results from several factors: less familiarity with AI technology, a higher level of trust in video content ("if you can see it on film, it must be true"), and less exposure to discussions about digital manipulation on social media.
Bank Pocztowy warned in its 2025 report that fake ads featuring celebrity deepfakes have become one of the main attack vectors against older customers. The pattern looks like this: the victim sees an ad online in which a well-known person encourages investing. They click the link and land on a professional-looking page. They register, deposit an initial amount | usually small, around $130-260. Then an "investment advisor" calls, offering to help "manage the portfolio" and pushing for further deposits. The money disappears for good.
Fake investments | 89% fraudulent domains
Fake investment platforms are currently one of the biggest financial threats to Poles | not just older adults, though they lose disproportionately large amounts. CSIRT KNF, the computer security incident response team at the Polish Financial Supervision Authority, published alarming data in 2024: as many as 89% of investment domains reported to the team turned out to be fake.
Fraudulent investment platforms have become dramatically more professional in recent years. Today's fake investment site looks like the website of a reputable broker | with charts, tables, a client panel, and even a mobile app. Some of these platforms generate fake "profits" on the victim's account to encourage further deposits. The older adult sees their "investment" growing and feels confident | not knowing the numbers on screen are fictional.
The mechanism used by fake platforms is usually multi-stage:
Stage 1: The ad. The victim comes across an ad on social media, in a search engine, or on a news site. The ad promises high returns, often using celebrity deepfakes. Headlines read: "Retiree from Katowice earned 50,000 zloty in a week" or "The new investment method banks don't want you to know about".
Stage 2: Registration. The victim lands on the professional-looking site and provides their details | name, phone number, email. This is the key moment: from now on, scammers have direct contact.
Stage 3: The first call. Within a few minutes, an "advisor" calls, speaking fluent Polish (often from a call centre abroad), explains "how the platform works" and encourages a first, small deposit. "Just $250 to start".
Stage 4: Building trust. "Profits" start appearing on the fake account. The advisor calls regularly, congratulates the victim on their "good decisions", and suggests increasing deposits. Some platforms even let the victim withdraw a small amount | so they believe the money is "safe".
Stage 5: Escalation. Scammers push for ever-larger deposits. Victims take out loans, borrow from family, withdraw life savings. According to CSIRT KNF data, the average loss reaches tens of thousands of zloty, but cases of losing several hundred thousand do occur.
Stage 6: Disappearance. When the victim tries to withdraw money, the platform demands a "withdrawal fee", a "profit tax", or "identity verification". Eventually the site vanishes, the phones go silent, and the money is gone.
How to defend yourself
The situation sounds grim, but it's not hopeless. There are concrete, practical rules that significantly reduce the risk of falling victim to fraud in the age of AI. Here are the five most important:
Rule 1: Verify identity separately
If someone claiming to be a family member calls asking for money | hang up and call that person on the number you have saved in your phone. Don't call back the number the call came from. Don't trust the voice | even if it sounds familiar. Agree on a family safety password that only you know, and use it in emergencies.
Rule 2: Don't trust celebrity investment ads
No well-known person advertises "secret investment methods" on social media. If you see such an ad | it's almost certainly a scam. Check the financial regulator's website to see whether the platform holds a licence. If it doesn't | don't deposit a single dollar.
Rule 3: Apply the 24-hour rule
Scammers always push for immediate action. "The offer expires in an hour", "You must deposit today", "There's no time to think". Adopt this rule: never make a financial decision under time pressure. Wait at least 24 hours. Talk to someone close to you. Most scams won't survive this test.
Rule 4: Check the source of information
Before you believe a message, ad, or phone call | verify the information through an independent source. Call your bank's official helpline (the number on your card, not from a text). Check your national cybersecurity team's website. Type the platform's name into a search engine along with the word "scam" | this is often enough to find warnings from other people.
Rule 5: Limit public information
The less information you share about yourself online, the harder it is for scammers to prepare a personalised attack. Set your social media profiles to private. Don't post voice or video recordings in open groups. Don't give your phone number to websites you don't fully trust. Any information you share publicly can be used against you or your loved ones.
At ODYSSEY, we teach all of this
AI-related threats are one of the key topics of the ODYSSEY programme. In our online course | specifically in Module 5 | we go into detail on how deepfakes, voice cloning, and fake investment platforms work. We show real examples, teach how to recognise warning signs, and practise applying safety rules.
Video 8 in our educational series is entirely devoted to AI-powered scams. In it, we show what the voice cloning process looks like, how a deepfake is made, and how to tell a fake investment platform from a real one. The video is available for free, with subtitles in three languages.
At our in-person workshops | launching in July 2026 in Kielce and Budapest | participants will be able to see with their own eyes how easy it is to create a deepfake, and practise reacting to suspicious calls in a safe setting.
Because knowledge is the best defence. Scammers count on their victims not knowing what's possible. The more you know about the tools they use | the harder it is to fool you.