Socure Adds Selfie Reverification with Deepfake Detection to its Industry Leading Document Verification Solution

Solution Powers Trust and Risk Decisions Across the Digital Economy with Near-Perfect 99.9% True Match Rate in Instant Identity Reverification

Socure, the leading provider of artificial intelligence for digital identity verification, sanctions screening, and fraud prevention, announced Selfie Reverification – a way to validate return consumers online with a simple selfie from any device in under 2 seconds.

The Selfie Reverification capability matches an incoming selfie with the previously verified ID headshot photo submitted during onboarding with a 99.9% true match rate. The scan also detects signs of a deepfake, as well as age discrepancy between the photo and the credential, to ensure the user is the rightful account holder.

Socure’s Selfie Reverification builds on the company’s market-leading Document Verification (DocV) product, which captures and validates photo IDs at onboarding with a 99% first-time selfie photo capture success rate, 98% accuracy in liveness detection, and a true acceptance rate of 98.2%.

With identity attacks at an all-time high, once validated good identities can go risky in an instant. Selfie Reverification allows simple, fast and accurate trust decisions particularly important at the riskiest of moments – a high-value banking transaction, validating the identity of a babysitter-for-hire in an online marketplace, verifying age for alcohol deliveries. The use cases for critical identity trust re-verification are almost endless across the digital economy and throughout the customer journey.

Marketing Technology News: MarTech Interview with Carlos Aragon, Senior Director, Product Marketing @ Vonage

As importantly, Selfie Reverification has integrated ‘liveness’ detection technology which ensures the person taking the selfie is actually present. This stops the most sophisticated spoofing techniques, including face masks, image-of-an-image and deepfakes.

“Identity verification isn’t a one-time event. As consumers interact with an online service over time, their risk profile can change. That’s why it’s important to determine you are still who you say you are, without going through the full verification process again,” said Pablo Abreu, Chief Product and Analytics Officer at Socure. “Selfie Reverification works seamlessly with our document verification solution guiding the user with real-time feedback capturing high-quality selfies. Simultaneously, it conducts Level 2 NIST PAD liveness checks and analyzes facial landmarks using our proprietary technology to prevent spoofing and stop deepfake selfie attacks without disrupting the user experience.”

Selfie Reverification enables organizations across verticals including financial services, online gaming, eCommerce, the sharing economy and the public sector to ensure a positive customer experience, while mitigating fraud, and lowering opex costs due to a reduction in manual reviews. Across these sectors, Selfie Reverification is a ‘Swiss Army Knife’ of rapid risk assessment across use cases and verticals.

Powerful Use Cases Include:

Account Takeover Prevention

• Selfie Reverification can be an integral part of account takeover prevention by triggering a reverification when unusual account changes are made that can allow a fraudster to gain control of the account.

Protection of High Risk Transactions

• Large financial transactions and wire transfers pose high risk as they can be initiated by someone other than the account owner. Selfie Reverification can be used as an additional – though frictionless – security measure to ensure that only the authorized user can perform these high-risk activities, protecting against fraud and saving consumers from devastating financial losses.

Account Lock-Out No More

• When users are locked out of their accounts or need to recover access due to suspected fraud, Selfie Reverification provides recovery in an instant with a rapid selfie check. Socure’s high speed capture and accurate matching reduce the customer friction and operational costs typically resulting from traditional account recovery methods.

Any Time ID is Necessary for a Quick Nod ‘Yes’

• From rideshare businesses which must re-verify drivers’ identities periodically throughout a shift to rapid loyalty member hotel check-in, Selfie Reverification can be leveraged with ease for any user revalidation dependent on documentary proof at re-validation.

Protect the Integrity of Government Services Delivery

• Selfie Reverification provides an added layer of security to protect returning users to ensure government benefits and services are provided to the true constituent.

Marketing Technology News: Mastering First-Party Data Part 1: 5 Tactics that will take you from zero to hero

How it Works:

User Action and Real-time Feedback: The user is prompted to take a selfie. During this process, they receive real-time feedback on positioning, angle, and lighting, ensuring the captured selfie is of high quality resulting in a decision in under 2 seconds.

Liveness Check: The selfie undergoes a liveness check that is Level 2 NIST PAD compliant to prevent fraudsters from spoofing the submissions through different presentation attacks, like images of images, or sophisticated masks. Adding to the liveness is Socure’s injection attack detection, which ensures a fraudster has not interjected a session with a false or altered credential. These checks are passive, so they do not disrupt the user experience or ask the user to do additional challenges, such as speaking numbers shown on the screen.

Deepfake Selfie Detector:  The Socure data science team generated hundreds of thousands of carefully curated deepfake samples using more than 20 different AI generators to ensure a data set consisting of a very realistic diversity of body shapes, ages, skin tones, postures, lighting conditions, backgrounds and ethnicities. These deepfakes train Socure’s selfie models to thwart spoofing attacks with incredible precision. Biometric analytics assess >80 facial landmarks, such as eye distance, nose width, jawline contours, and emotional expression, to create and verify a facial map to ensure accurate matching.