by Lenny Gusel
3 minutes • • December 16, 2024

Major Regulatory Alerts About AI and Deepfake Fraud Signal Significant Challenge to 2025 Digital Account Growth

Illustration of laptop screen with hands typing - two faces appear side by side on screen. Used to demonstrate the threat of Deepfake fraud

We’ve recently seen alerts from the FBI, the Financial Crimes Enforcement Network (FinCEN), and the New York Department of Financial Services about the use of AI and deepfakes against consumers and banks. Everyone’s paying attention, but the message to financial institutions is: growth, especially online, will suffer because of frustrations and limitations with legacy fraud controls. 

For years we’ve seen attacks against large banks. More recently these attacks have used AI and GenAI to cripple defenses. The new federal alerts aren’t news to the big banks; they’re somewhat capable of shifting with the fraud landscape. These alerts should send everyone else into serious planning mode. 2025 will prove to be a profitable year for criminals using AI, especially if fraud prevention strategies at small to mid-sized banks don’t adapt quickly. 

AI Deepfake Fraud vs. DocV: AI Will Win

Federal agencies have noticed a spike in suspicious activity reports highlighting the use of deepfake technology to manipulate driver’s licenses and other identity documents online. Cybercriminals are leveraging GenAI to create hyper-realistic fake identities that can slip through traditional verification processes like Document Verification or DocV. 

The best way to fight AI is with AI.

DocV is the way financial institutions verify a person is who they say they are by analyzing a government ID and matching it with facial biometrics. DocV was thought to be the silver bullet—high friction, high cost, but offering high assurance against fraud. It was the last line of defense before approving an account application.

 

The Path to Clear Customer Onboarding

Stop Fraudsters From Opening Bank Accounts With Synthetic IDs Fraudsters are using GenAI to create fake identities faster than ever. […]

Learn More

But there is no silver bullet to prevent fraud: DocV has been riddled with holes since day one. Cybercriminals have consistently found innovative ways to exploit these systems. Now, with the advent of advanced AI, they’re not just beating DocV—they’re doing it at scale.

What This Means for the US Fraud Landscape

The fraud game has changed, and so must the defenses. Multi-layered fraud prevention strategies at account opening aren’t just important—they’re paramount. FIs need to implement layered controls for DocV, ensuring that if one layer is compromised, others stand ready to block fraudulent attempts. But the key is to do this without causing additional friction for the customer.

Moreover, integrating workflow and data orchestration analytics at every decision point in the account opening funnel is no longer a luxury, it’s a necessity. By analyzing and cross-referencing data in real-time, institutions can detect anomalies and patterns indicative of fraud, enabling them to act swiftly. 

Redefine Fraud Prevention—Reputations are on the Line 

While the rise in deepfake fraud presents significant challenges, it also offers an opportunity. Enhancing KYC processes with advanced analytics could be a game-changer for the Bank Secrecy Act (BSA) and anti-money laundering efforts. By augmenting traditional methods with AI-driven analytics, like deep learning, FIs can better detect and prevent not just deepfake-related fraud but a broader spectrum of illicit activities. 

These federal alerts shouldn’t be taken lightly. The financial industry, especially those in mid-sized to smaller-sized, stand at a critical juncture where traditional defenses won’t work against sophisticated AI-driven fraud. The best way to fight AI is with AI, monitoring transactions in real-time and enhanced biometric monitoring. Sticking to a human element in this process will prove unsuccessful, AI can fight at scale, with speed.

It’s time for financial institutions to reassess and reinforce their defenses, embracing a multi-layered, analytics-driven approach to fraud prevention. Enhancing these processes doesn’t have to mean increased frustration for the customer account opening, done the right way, it can be seamless and be a selling point.

All expertise and insights are from human Feedzians, but we may leverage AI to enhance phrasing or efficiency. Welcome to the future.

Page printed in January 20, 2025. Plase see https://www.feedzai.com/blog/fbi-issues-ai-and-deepfake-alert for the latest version.