There was a time when a video was considered evidence.
If someone appeared on screen saying something, we assumed they had actually said it.
Artificial intelligence has changed that.
A convincing video can now be created showing someone saying words they never spoke. A voice can be replicated. A face can be manipulated. A photograph can be turned into something that looks entirely authentic.
For South Africans already dealing with sophisticated phishing, impersonation and investment scams, this creates a serious new problem.
Recent reporting citing identity verification company Sumsub found that deepfake incidents in South Africa had surged by 269% year on year, even while overall identity fraud declined. The pattern suggests that criminals may be carrying out fewer attacks but making them more sophisticated and targeted.
That matters because deepfakes attack one of our most basic instincts.
We believe what we see.
Why Deepfakes Are So Effective
A traditional scam might contain clues.
A suspicious email address.
Poor spelling.
An obviously fake photograph.
An unusual phone number.
Deepfakes can remove some of those clues.
A scammer can use the face of a celebrity to promote an investment.
The voice of a business executive can be replicated.
A familiar public figure can appear to recommend a product.
A family member’s face can potentially be manipulated into a video.
The technology is particularly powerful because it does not ask us to trust a stranger.
It borrows the identity of someone we already trust.
Investment Scams Are a Natural Fit
Investment scams are particularly well suited to deepfake technology.
People want reassurance before handing over their money.
A familiar financial expert can provide that reassurance.
A respected businessperson can create authority.
A celebrity can create excitement.
The scammer does not have to build credibility.
They borrow it.
This is why consumers should never make an investment decision simply because a famous person appears to endorse it in a video.
The endorsement itself must be verified independently.
Deepfakes Are Not Just About Celebrities
It is easy to think that deepfake scams only affect famous people.
That is not the case.
Ordinary people are also vulnerable.
Imagine receiving a video from someone who appears to be a family member.
Imagine an employee receiving a video call from someone who appears to be the company’s CEO.
Imagine a customer seeing a video of a financial adviser explaining why money needs to be transferred urgently.
As technology improves, these scenarios become increasingly plausible.
That means our verification habits have to change.
Seeing Is No Longer Enough
This is perhaps the most important lesson of the deepfake era.
Seeing something is no longer the same as verifying it.
A video can be fake.
A voice can be fake.
A photograph can be fake.
Even a live interaction can potentially be manipulated.
The answer is not to distrust everything.
It is to use more than one source of evidence.
If a family member sends an unusual request, call them using a number you already have.
If a public figure appears to promote an investment, check their official website and verified social media accounts.
If an executive requests an unusual payment, confirm through an independent channel.
Do not let the video become the verification.
The Emotional Trap
Deepfake scams work particularly well when combined with emotion.
Fear.
Excitement.
Greed.
Urgency.
A video saying you have been selected for an exclusive investment opportunity creates excitement.
A video warning that your account is at risk creates fear.
A video supposedly from someone you love asking for urgent help creates emotional pressure.
The stronger the emotion, the more important it becomes to slow down.
What Should You Look For?
There are technical clues that can sometimes indicate a deepfake, such as unnatural facial movement, strange lighting, inconsistent expressions or unusual lip synchronisation.
But relying entirely on spotting technical imperfections is not a good strategy.
AI-generated content is improving quickly.
The safer approach is behavioural.
Ask:
Why am I seeing this?
Who sent it?
What are they asking me to do?
Does the request make sense?
Can I verify it independently?
Is there pressure to act immediately?
These questions remain useful even when the technology becomes almost impossible to distinguish from reality.
A Final Thought
Artificial intelligence is not the enemy.
It is an extraordinarily useful technology that will continue to transform the way we work and communicate.
But every powerful tool can be misused.
Deepfakes have given scammers a new way to manufacture trust.
Scammers do not need to break into your life. They wait for a trusted person, platform or institution to open the door for them.
Increasingly, they can manufacture that trusted person.
That is why the most valuable skill in the deepfake era may not be learning how to spot a fake face.
It may be learning how to verify information without relying on appearances.
Pause.
Check another source.
Make the phone call.
Visit the official website.
Ask the question.
Because when seeing is no longer believing, verification becomes the new proof.