Facing the ID Authentication Deepfake Challenge
- Roy Urrico
- 3 hours ago
- 5 min read
By Roy Urrico
Deepfakes are now one of the fastest-growing fraud vectors worldwide with millions predicted to appear in fraud attempts in 2026, according to new analysis from Atlanta-based LexisNexis Risk Solutions, Unmask Deepfakes and Forged Documents with the Power of AI. With more than 100 billion digital identity checks expected globally this year 1 in every 100 failed identity checks now involve a deepfake -- an AI-generated liveness video, image, audio recording or document, that seems real.
The numbers point to a fraud vector accelerating faster than many unprepared organizations can handle, claimed the report. For example:
Deepfake-related attacks are up 180% year-over-year.
Passports, driver's licenses, and national IDs are the most frequently faked documents, with U.S., UK, German, and French documents most targeted.
One in 11 new account creations in 2025 was a fraud attempt, and nearly 20% of reported fraud involved unauthorized account access. One in every 11 new account creations in 2025 was a fraud attack and almost a fifth of all reported fraud involved unauthorized access of customer accounts, according to LexisNexis Risk Solutions’ latest Cybercrime Report.
Juniper Research estimated the number of digital ID verification checks carried out globally will reach 100.4 billion in 2026, up 16% year on year.
“Deepfakes vastly complicate digital identity verification. Protecting against this surge of attacks requires a solid line of defense incorporating end-to-end capture, fraud analysis and liveness checks.” said Kimberly Sutherland, Global Head of Fraud and Identity at LexisNexis Risk Solutions. “Even the smallest gap in your defenses is like an open window that a fraudster can climb through.”
“Businesses of all types are facing a wide range of fraud attempts and scams of increasing sophistication that identify weaknesses in business processes to exploit. Organizations may attribute less than 10% of fraud attacks to true identity theft as they are often experiencing surges in synthetic identity fraud and scam activity, especially financial institutions like banks and credit unions,” Sutherland told Finopotamus.
The Deepfake Threat
Deepfakes rarely fail on one obvious flaw the way physical forgeries do, explained the analysis. Instead, they slip through on a combination of subtle defects, in document structure, holograms, microtext, facial micro-expressions, and light reflection, that are nearly impossible to catch through manual review alone.
Fraudsters are leveraging generative AI to forge identities at scale— using deepfake images, video, audio and falsified documents to bypass traditional security protocols, according to the analysis. “These sophisticated schemes can strike at every stage: from account opening to high-value transactions and re-authentication—posing a growing threat to businesses around the world.”
When combined with imitators and tools that manipulate device and session metadata, “bad actors can infiltrate systems, steal sensitive data and drain assets across the global economy,” the report continued. To make it more severe, the increase of cheap, easily available generative AI tooling accelerates this trend—"enabling criminals to launch faster, more convincing attacks with unprecedented reach and scale.”
Bad actors use deepfakes to bypass identity checks and create new accounts — such as banking accounts — or take control of existing user accounts to make unauthorized payments, withdrawals and online purchases, launder the proceeds of crime, or abuse new customer bonus incentives.
As deepfakes become more realistic, identity checks need to be capable of spotting nuanced flaws in document security features and closely examine facial expression and skin tone. “Highly realistic deepfakes call for forensic examination of hundreds of security features: document structure, image integrity, holograms, etching and microtext. Deepfakes typically fail on several minor flaws, as opposed to physical forgeries that fail on one major issue, but they are not easy to spot with the human eye during manual checks,” Sutherland said. “The same goes for deepfake images and videos. Checks need to assess micro movements in facial muscles, analyses light reflection and detect image manipulation and injection tactics.”
Protecting Against ID Fraud
“As digital identity verification evolves, the core requirements shift toward the technical ability to integrate multiple trust signals into a coherent system architecture,” Shane O’Sullivan, research analyst at Juniper Research, noted. “Effective solutions depend on the coordination of document authentication, biometric liveness detection and real-time risk analysis within a single workflow. Increasingly, fraud detection system success is defined by how well it can detect advanced threats such as synthetic identities and deepfakes while maintaining interoperability across standards and minimizing latency and user friction.”
To combat AI-generated deepfakes and forged documents, businesses must meet the challenge head-on— by utilizing the power of AI ‘for the good,’ advised LexisNexis Risk Solutions in its breakdown. “This means adopting scalable, adaptable solutions that leverage the latest AI innovations to defend against emerging fraud tactics. AI-powered detection models can process massive volumes of data at high speed, continuously learning and evolving to identify new forgery techniques. These specialized systems are trained to spot subtle, often imperceptible, signs of manipulation that humans cannot spot—making them essential in the fight against generative AI-driven fraud.”
LexisNexis Risk Solutions offers IDVerse, which uses AI to “help businesses authenticate end users while disrupting deepfakes and forged documents from infiltrating their systems—quickly, more accurately and at scale,” said the report.
How Deepfakes Impact Credit Unions
The increasing availability of tools fraudsters can use to automate and improve their fraud attempts further compounds the problem, explained Sutherland. She noted, recent studies have shown that as high as 60% of attempted fraud involves digitally generated or manipulated identity data, documents or images. “Fraudsters are leveraging AI-powered ‘deepfake-as-a-service’ style websites where tools are available that make it faster and cheaper to execute these types of attacks with less experience and more precision.”
In addition, AI agents are fundamentally changing the scale, speed and operational consistency of fraud attacks, Sutherlkand pointed out. "Rarely can these types of attacks be identified by a human and they are increasingly able to deceive traditional verification methods."
Deepfake detection should operate as part of a broader identity, fraud and trust orchestration architecture rather than as an isolated point solution, Sutherland advised. “Synthetic identity and deepfake signals should be correlated with device and behavioral intelligence, document verification and authentication systems for the most effective risk decisioning across the full user lifecycle. “
Sutherland added, “Synthetic identity and deepfake media generation capabilities are evolving rapidly and static detection models can become operationally stale within months. Enterprise integrations should support continuous model deployment, rapid policy updates, fraud feedback loops and dynamic workflows without requiring major application or infrastructure redesign.”
There is no silver bullet, single technique or method to solve for the challenges of deepfakes attacks, warned Sutherland. “Organizations must deploy layered, explainable AI powered models and policy-driven risk decisioning processes with advanced detection capabilities to avoid struggling to keep pace with the rate of advancement in generative AI systems. The key for top performing organizations will be to protect against these evolving fraud attacks while prioritizing accessibility, usability and inclusiveness for a low-friction and equitable user experience.”