membership

What Are Deepfakes? Understanding the AI Tech Behind the Screen

Claire Grulick
Claire Grulick 5 Min Read
Businessman placing artificial face over real face symbolizing deepfake identity manipulation

Article overview

  • What are deepfakes? AI-generated videos or images that realistically alter a person's appearance or voice.

  • What are the risks? Deepfakes can be used to spread misinformation and threaten online safety.

  • What will I learn? This guide covers how deepfakes work, their legality and how to spot them.

  • How can I stay safe? Learn how to use tools like ProtectMyID to help protect your online identity.


Have you ever seen a video of a celebrity saying something completely out of character, only to realize later it wasn't them at all? Or maybe you've stumbled upon a clip of a politician giving a speech they never actually made. If so, you've likely encountered a deepfake.

As artificial intelligence (AI) continues to evolve at a breakneck pace, the line between reality and digital fabrication is becoming increasingly blurred. While this technology offers exciting possibilities for entertainment and education, it also brings a wave of new challenges regarding privacy, misinformation and online security.

In this guide, we'll break down exactly what deepfakes are, how the technology works and the different ways they are being used today. Most importantly, we'll cover practical steps you can take to spot them and protect your digital identity in an era where seeing isn't always believing.

stay protected everywhere...

Secure your digital and real-world identity

Don't let deepfakes or data breaches compromise your security. Join AAA today for expert identity theft protection, 24/7 monitoring and the peace of mind you deserve.

Join Today!

What are deepfakes?

A "deepfake" is a piece of synthetic media—usually a video, audio recording or image—in which a person's likeness is replaced with someone else's using artificial intelligence. The term itself is a blend of "deep learning" (a subset of AI) and "fake."

While photo editing has existed for decades, deepfakes take manipulation to a new level. Instead of manually airbrushing a photo, computers analyze thousands of images of a target person to learn exactly how their face moves, how they smile and how their voice sounds. The result is often a hyper-realistic forgery that can fool even the most observant viewers.

The potential impact of this technology is massive. On one hand, it allows for incredible creativity in filmmaking and art. On the other, the increasing prevalence and sophistication of deepfakes pose serious risks. They can be used to spread misinformation, damage reputations or commit fraud, making it essential for everyone to understand how this digital deception works.

Want to stay one step ahead of digital deception? Learn how to spot and avoid common AI-powered scams to protect yourself and your information.

Learn More
Example of AI Face swap or deepfake technology

How do deepfakes work?

At the heart of every deepfake is a complex form of artificial intelligence called deep learning. To understand how to deepfake myself or someone else (conceptually speaking), you have to look at the process of "training" a computer.

The role of AI and machine learning

Deepfakes use a special kind of computer brain called a "neural network" to learn. To make these videos look so real, two different AI programs play a game against each other. This is called a GAN.

Think of it like an artist and a judge:

  • The artist (the generator): This AI’s job is to create a fake picture or video from scratch.

  • The judge (the discriminator): This AI looks at the work and tries to spot if it’s a fake or a real photo.

At first, the Artist is bad at its job. But every time the Judge finds a mistake, the Artist learns and fixes it.

This loop continues millions of times until the fake is nearly indistinguishable from reality.

The creation process

Creating a convincing deepfake generally involves three main steps:

  1. Data collection: The creator gathers a massive dataset of the target person (the face they want to insert) and the source person (the body or face acting out the scene). This involves collecting hundreds or thousands of photos and videos from different angles and lighting conditions.
  2. Training: The AI analyzes the facial features of both people to understand the geometry of their expressions. It learns how the target's face looks when they blink, talk or laugh.
  3. Synthesis: The AI superimposes the target's face onto the source's body frame by frame. It adjusts skin tone, lighting and movement to ensure a seamless blend.
Man with laptop uses AI-based deepfake detection system to analyze manipulated facial data with high confidence

Types and examples of deepfakes

Deepfakes aren't just limited to swapping faces in movies. The technology is being applied in various sectors, ranging from harmless entertainment to malicious deception.

Deepfakes on celebrities

This is perhaps the most well-known category. You might have seen videos of Tom Cruise doing magic tricks on TikTok or classic actors "resurrected" for modern commercials. While many celebrity deepfakes are created for satire or entertainment, they raise questions about ownership of one's image. The ability to make a famous person say or do anything carries significant risk if used to damage their reputation.

Political deepfakes

Deepfakes have the potential to disrupt democratic processes. By creating realistic videos of politicians making controversial statements or engaging in illicit activities, bad actors can spread misinformation rapidly. In a volatile political climate, a well-timed deepfake could sway public opinion or incite unrest before the truth comes to light.

Deepfakes on children

A concerning trend involves the use of deepfake technology on minors. Parents often share photos of their children online without realizing that these images can be scraped to create deepfakes. This can lead to cyberbullying or other harmful scenarios where a child's likeness is manipulated. It serves as a stark reminder to be mindful of the digital footprint we create for our families.

Technology is an ever-changing landscape and the innocent act of posting your kids online could lead to irreparable damage. Here’s what you need to know about your children and their relation to the Internet. 

Learn More

Commercial deepfakes

On the other hand, businesses are using deepfakes for personalized marketing and customer service. For example, a company might use AI to dub a video into multiple languages, adjusting the speaker's lip movements to match the new audio perfectly. This allows for seamless global communication without the need for expensive reshoots.

Legal and ethical implications

As the technology advances, the legal system is playing catch-up.

Are deepfakes legal?

The legality of deepfakes is complicated and varies by location. In general, creating a deepfake for satire or artistic expression is often protected under free speech laws. However, using deepfakes for fraud, defamation or non-consensual pornography is illegal in many jurisdictions. Governments worldwide are currently drafting new legislation to specifically address the malicious use of synthetic media.

Can you sue for a deepfake?

Yes, in many cases, you can sue. If a deepfake is used to defame you, invade your privacy or misuse your likeness for commercial gain without permission, you may have grounds for a lawsuit. The challenge often lies in identifying the anonymous creators behind the content and proving the specific damages caused by the video.

Caring mom providing children's online privacy protection

How to protect yourself online

While you can't stop the technology from existing, you can get better at spotting it and protecting your own data.

Tips for identifying deepfakes

Even high-quality deepfakes often leave digital clues. When watching a video that seems suspicious, look for these tell-tale signs:

  • Unnatural eye movements: Does the person blink too little or too much? Do their eyes look vacant?

  • Lip sync issues: Does the audio match the mouth movements perfectly? Sometimes the lips might look blurry or robotic.

  • Lighting inconsistencies: Look at the shadows on the face compared to the rest of the scene. Deepfakes often struggle to match the lighting of the original video perfectly.

  • Skin tone mismatches: Check the edge of the face where it meets the hair or neck. You might see a slight difference in skin texture or color.

  • Glitching: Watch for flickering or jittering, especially when the person turns their head or moves quickly.

Best practices for online security

Protecting your digital identity is your first line of defense against becoming a target for deepfakes or identity theft.

  • Lock down privacy settings: Review the privacy settings on your social media accounts. Limit who can see your photos and videos to friends and family only.

  • Use strong, unique passwords: Never reuse passwords across accounts. Use a password manager to generate complex credentials.

  • Limit public sharing: Think twice before posting high-resolution images of your face or your children's faces on public forums. The less data available, the harder it is for AI to mimic you.

  • Enable two-factor authentication (2FA): Add an extra layer of security to your email and social media accounts to prevent unauthorized access.

Watch our expert-led webcast to learn how to spot identity scams and keep your personal data safe in the age of AI.

Watch the Webcast

Safeguard your identity with Experian

In a world where digital identities are increasingly vulnerable, having a robust protection plan is essential. Identity theft can happen to anyone and the consequences can be financially and emotionally draining.

As a AAA Member, you don't have to face this challenge alone. AAA offers ProtectMyID®, a powerful identity protection service provided by Experian®, a brand you already trust.

Free protection for members

Did you know that Essential protection is included for free with your AAA Membership? This plan provides you with daily credit monitoring from Experian, lost wallet protection and fraud resolution support.

How ProtectMyID helps

ProtectMyID helps safeguard you against identity theft by:

  • Detecting threats: It spots early signs of fraud with tools like Dark Web Surveillance (SSN and email monitoring) and Real-Time Identity Alerts.

  • Resolving issues: If you do become a victim, dedicated Fraud Resolution Agents are available to assist you in restoring your identity.

  • Insuring you: Premier members receive up to $10,000 in identity theft insurance, while upgraded plans offer even higher coverage limits.

Whether you stick with the free Essential plan or upgrade to the Deluxe or Complete plans for features like social media monitoring and child identity monitoring, ProtectMyID offers you the peace of mind you deserve.

Stay alert and stay safe

Deepfakes are a fascinating, yet intimidating, example of how powerful AI has become. While we can expect to see more of this technology in movies and social media, the potential for misuse is real.

By understanding how deepfakes work and staying vigilant about the content you consume, you can avoid falling for digital tricks.

stay protected everywhere...

Secure your digital and real-world identity

Don't let deepfakes or data breaches compromise your security. Join AAA today for expert identity theft protection, 24/7 monitoring and the peace of mind you deserve.

Join Today!