What Are Deepfakes?
A deepfake is a type of media - video, audio, or image that has been digitally altered using Artificial Intelligence (AI) to make it look and sound real. The word comes from “deep learning” (a form of AI) and “fake.”
Deepfakes are not just “fun internet tricks.” While some are made for entertainment, many are created for fraud, scams, or spreading false information.
A video can show a politician giving a speech they never made.
A phone call can sound like your boss asking you to transfer money.
A family member’s voice can be copied to request emergency help.
At first glance, these seem very real - and that is what makes them dangerous.
Why Should You Care About Deepfakes?
Deepfakes are becoming more common and more convincing. Criminals use them because they exploit trust and emotion.
Here are some real-world risks:
Phone scams: A deepfake voice asks you for a quick money transfer, pretending to be your boss or loved one.
Financial fraud: A fake video or audio convinces you to give away money or sensitive data.
Disinformation: Fake videos or images spread online, designed to manipulate opinions or damage reputations.
Identity theft: Your face or voice could be stolen and used without your permission.
The main danger is that deepfakes target emotions - urgency, trust, fear, or authority to make you act without thinking.
Criteria to Identify Deepfakes
photo generated by GenAI
Even though deepfakes can look very real, they usually leave behind small errors or visual/audio clues. By knowing what to look for, you can protect yourself from being fooled.
1. Facial Inconsistencies
Blurry edges around the face or hairline.
Unnatural skin texture (too smooth or plastic-looking).
Lighting and shadows that don’t match the rest of the scene.
2. Eye and Mouth Movements
Eyes blink less often or blink unnaturally.
Lips may not fully match the words being spoken.
3. Object Interference (Masking)
When something passes in front of the face (like a hand or object), the deepfake may “glitch” or ghost.
4. Asymmetry
One side of the face may look different from the other.
Background objects (like glasses, earrings) may be distorted or mismatched.
5. Audio Red Flags (for voice/video deepfakes)
Robotic or flat tone.
Missing natural breathing sounds.
Background noise doesn’t fit the environment.
Tip: If you notice one or more of these signs, treat the content as suspicious and verify through other trusted channels before believing or acting.
How Can You Protect Yourself?
1. Pause Before You Act
If you receive an urgent request for money, sensitive information, or immediate action, stop and think.
Scammers want you to rush. A short pause can save you from being tricked.
2. Verify Through Trusted Channels
Call back using the official phone number (not the one given in the suspicious message).
Confirm with the real person through another method (e.g., email, face-to-face, secure chat).
3. Look for Visual and Audio Signs
Deepfakes often have small mistakes. Watch and listen carefully:
In videos: unnatural blinking, mismatched lip movements, strange lighting, or blurry edges.
In voices: robotic tone, missing natural breathing sounds, or unusual background noise.
4. Don’t Believe Everything Online
If you see a shocking or unbelievable video on social media, be cautious. It could be a manipulated deepfake designed to spread false information.
5. Report Suspicious Activity
If you suspect a deepfake is being used against you or your organization, report it immediately to customer support or the security team. Early reporting helps protect you and others.
Key Takeaways
Deepfakes are fake videos, images, or voices created with AI.
They can be used for scams, fraud, and spreading lies.
Always pause, verify, and double-check before taking action.
Remember: If something feels “off,” it probably is.