By Jackson Godwin. Cybersecurity Analyst & Penetration Tester.
I Called the Real Person
As soon as I ended the video call, I picked up my phone and dialled the person’s actual number.
They answered almost immediately.
Before I could even explain why I was calling, they sounded confused.
“Is everything okay?”
I asked a simple question.
“Didn’t you just video call me?”
There was a long pause.
Then they replied:
“No… I’ve been in a meeting for the past hour.”
At that moment, I realised something frightening.
The person I had just seen on my screen wasn’t them.
The Video Had Been Generated by AI
After describing what had happened, we both started researching similar scams.
What we found was alarming.
Cybercriminals are increasingly using artificial intelligence to create convincing impersonations.
By combining publicly available videos, photographs and audio recordings, attackers can generate fake voices and realistic-looking video calls.
This technology is commonly known as a deepfake.
While deepfake technology also has legitimate uses in entertainment, education and filmmaking, it can be misused for fraud and impersonation.
Why It Looked So Real
The scam worked because the criminals had copied details I already trusted.
The face.
The smile.
The voice.
The mannerisms.
Even small habits, like pausing before answering or laughing in a familiar way, had been imitated.
The attack wasn’t technically perfect.
But it was convincing enough to make me lower my guard.
Then I Remembered the Strange Moments
Looking back, several things hadn’t quite felt right.
The video occasionally froze.
The lips didn’t always match the words perfectly.
The lighting changed unexpectedly.
At the time, I blamed the internet connection.
Now I realised those tiny glitches may have been signs that something wasn’t genuine.
Not every deepfake will show obvious flaws, but unusual visual or audio inconsistencies can sometimes be clues.
The Criminal Didn’t Want Me to Verify
The biggest warning sign wasn’t the technology.
It was the behaviour.
Every time I suggested another way of confirming the request, the caller found an excuse.
“Don’t call me back.”
“My microphone isn’t working.”
“I’m using someone else’s phone.”
“Please hurry.”
Those excuses had one purpose.
To stop me from verifying their identity.
The Scam Relied on Emotion
The caller didn’t begin by asking for money.
Instead, they spent several minutes building trust.
We talked about familiar topics.
They mentioned real events.
They behaved like someone I genuinely knew.
Only after I felt comfortable did they introduce the emergency.
That sequence wasn’t accidental.
It was carefully planned.
The criminals understood that trust comes before persuasion.
AI Didn’t Steal the Money
One important lesson became clear.
Artificial intelligence wasn’t the real threat.
The real threat was social engineering.
The AI simply made the impersonation more believable.
The criminal still relied on:
- Creating urgency.
- Building trust.
- Preventing verification.
- Encouraging quick decisions.
Technology made the deception easier.
Human psychology made it effective.
One Habit Saved Me
Looking back, one simple decision protected me.
I refused to act until I confirmed the request through another trusted method.
That single phone call exposed the entire scam.
Had I relied only on what I saw during the video call, I might have transferred money to a criminal pretending to be someone I trusted.
In the final part of this article, I’ll explain the biggest warning signs of AI-powered video call scams, share practical ways to verify someone’s identity, and reveal the simple rule I now follow whenever a video call involves urgent requests or financial decisions.
Continue Reading: The AI Video Call That Fooled Everyone (Part 3)