Market Guide for Identity Verification
Gartner® Magic Quadrant™ for Identity Verification
Get the Report
X
Essential Guide

Advanced Deepfake Detection

What it is, why it matters and essential strategies to protect against sophisticated identity fraud tactics that can diminish trust and profits.
CHAPTER 1

New Era of Fraud: The Rise of AI & Deepfakes

What is a deepfake?

Deepfakes are highly realistic, digitally manipulated videos or images where a person's likeness, voice, or actions are artificially created or altered using deep learning techniques, especially with neural networks. These tools can generate highly convincing media that appears real but is, in fact, synthetic. Deepfakes often use techniques like generative adversarial networks (GANs) to blend and map one person’s facial expressions or voice onto another, creating the illusion that the subject is saying or doing things they never did.

GIF of faces interchanging showing a deepfake swap.

The alarming surge of deepfakes

60%

of consumers believe they could detect a deepfake, up from 52% in 2023. (Jumio)

53%

of crypto firms have encountered video deepfake fraud. (Regula)

200%

Presentation attacks are the most common attack vector, but injection attacks increased 200% in 2023. (Gartner)

deepfake icons

“The number of deepfakes detected worldwide in 2023 was 10 times the number detected in 2022. Gartner estimates the time to reach the early majority (i.e., more than 16% target market adoption) for deepfakes is one to three years because deepfakes go hand-in-hand with the GenAI advances that underpin their creation. This requires that identity verification vendors take a multipronged approach to safeguard against these rising deepfake attacks.”

2024 Gartner® Emerging Tech: The Impact of AI and Deepfakes on Identity Verification, Swati Rakheja, Akif Khan, 8 February 2024
Back to top
CHAPTER 2

Emerging Deepfake Tactics and the Tools Fraudsters Use

Deepfake videos are created in one of two ways. Fraudsters can use an original video source of the target, where the person is made to say and do things they never did, or they can swap the person's face onto a video of another individual, also known as a face swap.

image of two women photos. Woman on the left has short to mid brown hair. Woman on the right has same hair but different face.

Types of Deepfakes Used to Trick KYC

black and white image of two faces being morphed.

Face Morphs

This method merges two (or sometimes more) different faces into a morph, a new face that retains biometric characteristics of both at the same time.

black and white image of person holding piece of paper infront of their face with a headshot of a different person.

Synthetic Face

Face synthesis technology is capable of modeling a fake and nonexistent face from scratch leveraging GANS (which take advantage of two competing neural networks) and other specialized digital tools.

black and white image of split screening showing one half of face and another.

Face Manipulations

This method involves manipulation of facial attributes (e.g., hair, eye, color, skin texture). This technique is used to manipulate both photos and videos, and is often used with face swapping.

black and white image for voice cloning

Voice Cloning

The bad actor clones audio of a person's voice, creates a model based on the vocal patterns and uses that AI model to make the voice say anything the creator wants.

black and white image of CA ID.

Synthetic Identity Document

Deepfake technology is also being used to create convincing fake identity documents, such as driver’s licenses or passports. While these ID documents aren't necessarily a deepfake themselves, fraudsters use the same deepfake technologies to modify the face image on the ID.

60% Increase in Deepfake Tools

In 2023, there was a significant increase in the creation of AI deepfake tools, with over 60% being newly developed.

Companies Are Unprepared

80% of companies lack tools that detect deepfakes today.

Tools Used to Create Deepfakes

GAN neural network technology:

GAN neural network technology uses generator and discriminator algorithms to develop all deepfake content.

Natural language processing:

Natural language processing is used to create deepfake audio. NLP algorithms analyze the attributes of a target's speech and then generate original text using those attributes.

Convolutional neural networks:

Convolutional neural networks analyze patterns in visual data. CNNs are used for facial recognition and movement tracking.

High-performance computing:

High-performance computing delivers the significant power required for deepfakes.

Autoencoders:

Autoencoders are a neural network technology that identifies the relevant attributes of a target, such as facial expressions and body movements, and then imposes these attributes onto the source video.

video editing icon

Video editing software:

Video editing software isn't always AI-based, but it frequently integrates AI technologies to refine outputs and make adjustments that improve realism.

Back to top
CHAPTER 3

How Video Injection Works

What is Video Injection?

Video injection is a type of digital injection attack where fake data, such as AI-generated documents, photos or biometric images, is introduced into the data stream of an identity verification platform. This technique bypasses the use of a physical camera by leveraging emulators to simulate a hardware camera. As a result, it can circumvent the platform's ability to verify the presence of a real user in front of the camera.

GIF of video injection

How Injection Attacks Impact Identity Verification

Injection attacks can be problematic for identity verification systems, especially those that use video-based authentication. Here's why:
video injection icon graphic
  • The attacks can bypass facial recognition or liveness detection by mimicking a legitimate user.

  • The attacks can trick less sophisticated systems that fail to recognize the difference between genuine live input and injected video.

  • Attackers can impersonate legitimate users to open accounts or authorize financial transactions.

  • Successful attacks erode trust in the system, leading to customer dissatisfaction and even potential regulatory fines.
Back to top
CHAPTER 4

The Impact of Deepfakes on Identity Verification

Gartner estimates that by 2026, nearly one-third of enterprises will consider identity verification and authentication solutions unreliable due to AI-generated deepfakes.

Deepfakes can allow criminals to successfully bypass customer identity verification procedures used by financial institutions, thereby eliminating the need for living accomplices.

These accounts are later used for various criminal activities — mainly for direct financial fraud, as well as laundering profits from illegal operations. Regulators will likely tighten requirements for fully remote financial operations. It’s possible that the simplicity and convenience of online financial services, which we’ve already become accustomed to, will be threatened by AI.

Unfortunately, the problem doesn’t end there. As noted by experts, the widespread availability of AI tools for generating photo, video and audio content fundamentally undermines trust in digital interactions between people. The higher the quality of AI creations, the harder it becomes to believe what we see on our smartphones and computers.

Expand Your Knowledge

Combating Deepfakes: Is Your Business Doing Enough?

Watch the Webinar Replay
Back to top
CHAPTER 5

The Anecdote: Liveness Detection

What is liveness detection?

Liveness detection is a technique used by identity verification solution providers to ensure that the biometric sample (such as a fingerprint, face or voice) being presented is from a live person rather than a static image, video or other spoofing method (e.g., a deepfake). It is a crucial security feature in online verification and authentication solutions to prevent spoofing attacks, where someone might try to create an online account or gain unauthorized access by presenting a photo, mask or other fake representation of a person’s biometric trait.

GIF shows Jumio liveness process

Detection strategies

Active liveness detection

The user is prompted to perform a specific action, such as blinking, nodding or smiling. This method makes replicating a live person with a static image or mask difficult. However, this process can introduce unnecessary friction and lower overall conversions.

Passive liveness detection

This method analyzes subtle characteristics of live biometrics, such as blood flow patterns in fingerprints, depth information in facial scans, or muscle movements during voice recognition. Passive methods are often more seamless for the user but may require more sophisticated algorithms.

Semi-passive liveness detection

This method may not ask for a specific movement from the user. Still, it may project blinking lights or randomized colors onto a user's face or zoom a camera to detect depth information.

Texture analysis

This method is widely used in face liveness detection systems because it can differentiate between live and spoof facial samples.

Motion analysis

This method assesses the subject's movements to determine if it's a live person. It looks for natural motions and behavior patterns that a live human would produce.

Multimodal biometrics

This method combines voice and face biometrics, as well as voice and facial liveness detection. Learn more here.

image of tradeoff icon

The tradeoff between strong liveness and a smooth UX

Liveness detection is now table stakes for remote identity verification. Virtually every reputable identity verification solution has incorporated liveness detection into the new account onboarding process. However, the real differentiator lies in how effectively these solutions can detect advanced deepfakes while maintaining a seamless user experience.

Overly stringent or intrusive liveness checks, such as excessive prompts or repeated scanning attempts, can frustrate users and lead to onboarding abandonment. On the other hand, too lenient or simplistic checks may fail to detect sophisticated spoofing attempts, compromising security.

Striking the right balance requires leveraging advanced AI and biometric technology to deliver robust fraud detection without adding unnecessary friction. The ideal liveness detection system is fast, accurate and intuitive — providing clear instructions to users while working behind the scenes to identify spoofing attempts, such as video injection attacks or AI-generated images.

Ultimately, user trust hinges on a solution’s ability to provide both security and a smooth, hassle-free experience, ensuring customers feel protected without being inconvenienced.

Back to top
CHAPTER 6

Liveness Detection Use Cases

Financial Services

  • Coerced account opening prevention
  • Account takeover prevention
  • Unauthorized high-risk transactions
financial services graphic
gaming graphic

Gaming

  • Fake or duplicate account opening prevention
  • Age restrictions
  • Bonus abuse prevention

Shared Technology

  • Passengers and guest protection
  • Driver and host protection
  • Account takeover presentation
Shared technology graphic
Back to top
CHAPTER 7

Security Standards and Liveness Detection

security icon
Because deepfakes are evolving so quickly, there are not many ways to certify how effective an identity verification solution is in catching these types of attacks. Here’s a quick overview of what is available today regarding certifying bodies.

ISO/IEC 30107-3

An international standard that provides guidelines for testing and reporting on liveness detection in biometric systems. The standard is designed to help ensure that biometric systems can prevent and detect fraudulent activities, such as when someone tries to bypass the system using fake biometric samples. Here again, this standard focuses on presentation attacks versus deepfake detection.

iBeta

iBeta is a biometric security testing company that offers presentation attack detection (PAD) testing services in accordance with the ISO/IEC 30107-3 standard and in alignment with the ISO/IEC 30107-1 framework.

This standard is recognized globally, especially for its application in biometric authentication and identity verification solutions. It ensures technology is resilient to spoofing, aligned with data privacy requirements and capable of providing secure, real-time identity verification. NOTE: iBeta does not test whether an identity verification solution can detect deepfakes — it only has testing standards for presentation attack detection.

The National Voluntary Laboratory Accreditation Program (NVLAP)

Provides third-party accreditation to testing and calibration laboratories in response to legislative actions or requests from government agencies or private-sector organizations. NVLAP-accredited laboratories are assessed against the management and technical requirements published in the International Standard, ISO/IEC 17025:2017.

Back to top
CHAPTER 8

Why the Capture Channel Matters

To minimize risk while maximizing convenience for users, it’s crucial to choose the right capture channel for liveness detection. Different channels offer varying capabilities to ensure accurate verification while maintaining user-friendly experiences.

Channels like mobile are particularly well-suited for liveness detection due to their combination of advanced hardware, consistent configurations and seamless integration with biometric features. In contrast, web and API channels, while flexible, may present challenges such as inconsistent camera quality and vulnerabilities to fraud. Selecting the right channel involves balancing robust fraud prevention with user accessibility, and mobile often provides the ideal solution for high-stakes identity verification scenarios.

Learn more about which identity verification channel is right for you.
Expand Your Knowledge

7 Questions to Ask Your Liveness Detection Vendor

Watch the Webinar Replay
Back to top
CHAPTER 9

How Jumio Can Help

Watch a Quick Explanation of Jumio Liveness

Providing the reliability, adaptability and security businesses trust to stay protected.
Watch now

Next-Gen Liveness Detection

AI-driven technology
analyzing real-time user behavior to prevent deepfakes, masks and spoofing attempts.
ISO/IEC 30107-3 compliant
meeting rigorous industry standards for biometric security and fraud prevention.
Conforms to NIST/NVLAP testing standards
ensuring top-tier accuracy and reliability.
Detects sophisticated fraud
while staying compliant, delivering a secure and seamless customer experience.

Covering Industry Standard Checks

Built on ISO compliance standards, ensuring that our checks meet rigorous industry requirements for security, accuracy and reliability.

ID Image used as
Selfie
Paper
Printouts
Digital
Copy
Face
Masks

Beyond the Standard

Image Quality Checks

Incorporating advanced image quality checks to help ensure precise verification, detecting fraud with accuracy even in challenging environments.

Face Not Fully Visible
Is there a covering over the face?
Multiple People
Is there more than one person in the image?
No Face Present
Is there a nose, mouth and eyes?
Black and White Image
Is the image in color or black and white?

Deepfake Detection

Leveraging cutting-edge deepfake detection technology to identify even the most sophisticated synthetic fraud attempts, safeguarding the integrity of your identity verification process.

Synthetic Images
Face / Head Swaps

Injection and Replay Detection

Including robust camera and video injection detection, blocking fraudulent attempts at bypassing verification by using pre-recorded or manipulated content.

Additional Fraud Checks

Offering advanced fraud prevention features, including detecting sleeping individuals and manipulated selfies. Identifying inactive users and tampered images to help ensure only genuine, alert individuals pass verification.

svg
“Online platforms hold a critical duty to leverage cutting-edge detection measures like multimodal, biometric-based verification systems to fortify our defenses against deepfakes."
Daryl Huff, Vice President of Biometrics
Learn more about the world of biometric security.

Biometrics & Fraud Analytics: Stopping Fraud at the Source

Watch the Webinar Replay
Back to top

Get Started

Let a Jumio expert show you how easy it can be to integrate our automated solutions into your existing processes.
Request a Conversation
image of man with facial hair smiling wearing a suit.