A Relationship That Was Never Real
Maria matched with a profile on a dating site; the man messaged her constantly, even overnight. “It felt like there was a love spell that connected our minds,” she told AFP, asking that her name be withheld. He introduced himself on video calls as Sheikh Hamdan bin Mohammed bin Rashid Al Maktoum, known as Fazza. His lips matched his words, but his voice sounded slightly off — a mismatch voice-cloning tools still leave behind.
She sent 100,000 pesos ($1,625) for a marriage certificate and “royal membership card.” A second ask for 60,000 more broke the spell. She traced his deleted Facebook account to Nigeria. His likeness has been used before. AFP identified several Facebook groups impersonating the royal, some with thousands of followers. These groups direct users into WhatsApp or Telegram chats with the “prince.” A Change.org petition now urges his office to warn the public.
Same Script, Cheaper Repetition
Romance scams have worked the same way for decades: fake urgency, a fake identity, then a money request disguised as paperwork. The one thing that limited scammers was time and effort — a single person could only keep up a few fake identities before the lies started contradicting each other and the scam fell apart. AI changes that math: the same fake face and voice gets reused on a new target almost for free.
French authorities opened a similar case last year involving a cloned Brad Pitt persona that cost a woman 830,000 euros ($945,000), per AFP — a different face, same reusable setup. Cornell researcher David Rand told AFP he expects it to get harder to tell a real call from a fake one as the tools improve. He hasn’t seen that happen yet, but the pace of recent releases points that way.
What the FBI’s Numbers Actually Show
(AI Scam by The 2025 IC3 Annual Report)
The FBI’s 2025 IC3 report logged 22,364 complaints citing AI last year. Those complaints added up to nearly $893 million in reported losses. That volume shows how far this pattern has spread. But the number itself is looser than it sounds. The AI descriptor covers any complaint that mentions AI — not an audited count of losses AI actually caused. And the report doesn’t break out how much came from video deepfakes specifically, versus AI-written scripts or cloned voices used in other kinds of fraud.
Maria’s case shows why the real number is likely bigger. She filed no U.S. complaint. A scam that ran on the exact mechanism the FBI is tracking simply isn’t in its count.
Why Nobody’s Closing the Gap
Dating apps verify a phone number, not a face. Once chats move to WhatsApp, nothing checks the identity behind the call. Liveness checks and voice-artifact detection exist in theory, but running them on every call is expensive at scale. WhatsApp is also end-to-end encrypted by design, so even the platform can’t see call content to scan it — the same encryption that protects privacy blocks any check on who’s on the line.
Money explains why that gap stays open. A credit card fraud loss lands directly on the card company, so it pays to build detection. A dating app doesn’t lose money that way — the scam money moves through a separate bank transfer, after the damage is done. Dating apps make money by connecting people, not by verifying them. Nothing forces that to change unless a law, an insurance rule, or a big lawsuit makes the app pay too.
Maria wasn’t careless; she trusted someone who said what she wanted to hear. What’s different now is how many people that trust can be drained from at once.
FAQs
Does the FBI’s $893 million figure include cases like Maria’s?
No. It only counts U.S. complaints that reference AI, not a confirmed global tally — so hers sits outside it entirely.
Is it getting harder to spot a fake video call?
Yes, but not impossible yet — mismatched audio and lip movement, like the flicker Maria noticed, still catch some calls out.
Where do operations like this typically originate?
AFP linked some networks behind the Sheikh Hamdan impersonation to Nigeria, though scammers often route payments through other countries to stay hard to trace.
Why don’t dating apps just add identity verification to video calls?
Mostly cost and liability. Real-time detection is expensive and risks flagging real users as fake. And the loss from a scam lands on the victim, not the app — so there’s little financial pressure to build it.
Click Here For The Original Source.
