The What, How, and Why of Identity Document Verification

The rapid surge in digitalization has spurred new business models and tremendous growth for a wide array of businesses. For organizations like neobanks, two-sided marketplaces, telehealth, and alcohol delivery, this growth is fueled by the ability to efficiently onboard new customers and provide safe, secure ecosystems. Excelling in digital customer acquisition and trust and safety has become core to the bottom line for many emerging digital businesses.

Simultaneously, artificial intelligence has enabled fraudsters to scale their craft by creating sophisticated forged IDs, deepfakes, 3D masks, and more. The output is so convincing that human eyes often fail at detecting these fake artifacts.

Challenges arise when businesses that process IDs to onboard customers do so by applying manual reviews, implementing subpar technology to verify IDs, or not using any identity document verification method at all. Instead of quickly approving legitimate customers to support growth, these methods make customers wait to access services, cause errors that block good users because they appear risky, or subject innocent people to possible harm by allowing dangerous criminals to use their platform.

So how should identity document verification work in a digitally evolving world?

What is Identity Document Verification?

Document verification is the process of verifying the authenticity of a government-issued ID, including driver’s licenses, passports, and state or national identification cards.

Identity document verification addresses these questions:

1. Is this ID authentic?
2. Does the presented identity match the person on the ID?

With digital onboarding, applicants no longer need to be physically present. For even fast and secure onboarding, an ID verification could happen upfront, simultaneously complying with regulations requiring enhanced verification – such as age verification for restricted purchases – while simultaneously preventing fraud. In other cases, the ID and selfie ID verification check is further down the waterfall as step-up verification when other checks are inconclusive.

In either case, an effective image capture app allows the applicant to snap a picture of the front and back of an ID and then upload an optional selfie to biometrically check if it matches the photo on the ID – after which they should get a quick response, ideally within seconds.

How Does Digital Identity Document Verification Work?

The key to fast, accurate document verification in a digital world is to rely on robust, automated machine learning (ML), which conducts hundreds of validation checks on the ID and selfie in seconds versus an uncertain in-person glance during a manual review. The advanced image capture app proactively guides the applicant on how to take high-quality ID and selfie photos by eliminating blur, glare, and other issues to create the optimal circumstances for a good user to pass security checks on the first try.

Here are just some of the hundreds of automated checks that organizations should require from their document verification vendor:

• Face and orientation detection
• Edge detection and cropping
• Front-to-back ID data extraction and correlation
• Barcode forensics
• Alert list cross-check
• Accessibility
• Headshot integrity
• Selfie-to-ID photo match and liveness detection

ML-driven document verification depends on computer vision technology and correctly labeled datasets to attain high accuracy. The models are trained on this data and learn continually, compounding performance. When assessing the advantages of automated ML-powered document verification over manual processes, there are numerous benefits:

• More accurate forgery, tampering, and spoofing detection
• Adaptable to quickly scaling; no latency from volume spikes
• Stronger auto decisioning and fewer falsely rejected good users
• Faster onboarding and time to revenue
• More satisfying user experience

Types of Document Verification

There are a few types of document verification that help solve for a range of use cases across industries.

1. Document Only Use Case: When document verification is the only means of verifying an identity, then it must be a robust system with an effective process for weeding out bad actors. Keeping out fraud while identifying trusted users requires a robust understanding of identity, forensic analysis, and machine learning.
2. Document First Use Cases: When an ID check is used early in the verification waterfall, the solution needs to be as user-friendly as possible to reduce abandonment rates. When additional verification and authentication is required in the future, a document verification platform provides significantly more value when it can link to a robust identity graph and complete verification platform, ensuring a verified user has a continuous positive experience while repeat offenders are kept out of the loop system.
3. Step-Up Verification Use Case: When document verification is further down the waterfall when a KYC check is inclusive or to step-up fraud prevention, document verification is the last step before a transaction happens, and these are typically the riskiest transactions. Stopping the fraud requires a robust understanding of the identity presented earlier in the waterfall as well as an ID check with forensic analysis and adaptive machine learning.

Documents Used for Verification

To quickly and accurately confirm their identity, users provide ICAO-compliant travel documents and national ID cards.

Why is Document Verification Important?

In the introduction, we covered the growing markets and use cases for identity document verification, including the need to detect AI-generated fake IDs and selfies. Accurate, automated document verification aids companies in identifying anomalies to prevent fraudsters and criminals from entering the ecosystem. At the same time, it facilitates the auto-approvals of legitimate customers–which is why a fast and seamless user experience is critical.

Today’s consumers expect streamlined, low-friction onboarding and immediate account or service access. As digital document verification becomes more mainstream, a requirement for multiple selfies, head-turning video, or subsequent manual processing creates a frustrating user experience that creates delays and dropoff, which negatively impacts your brand’s reputation. At a time when two in three consumers report having abandoned a remote bank account application due to a poor experience, businesses must deliver seamless onboarding to meet users’ expectations.

Predictive DocV 3.0: Fast, Accurate, User-Friendly Document Verification

Socure’s Predictive Document Verification (DocV) 3.0 is the market-leading identity document verification solution with superior accuracy, speed, user experience, and fraud prediction performance.

Computer vision-driven image capture technology identifies blurry or incorrectly positioned images and automatically coaches users through the ID and selfie intake process to reduce friction, while selfie-to-ID-photo matching ensures that the applicant is who they say they are and passive, single-frame liveness detection thwarts spoofing attacks.

With a powerful forensics and data extraction engine, DocV 3.0 extracts hyper-accurate data such as OCRs, MRZs, and barcodes, with advanced image processing to maximize user experience while accurately flagging fraudulent documents. And all of this is done with a market-leading average response time under three seconds, with 93% accuracy in accepting good customers on the first try versus an industry standard of 65%.

Relying on ML fraud models trained on 400+ third-party data sources and leveraging Socure’s comprehensive identity graph of one billion known good and bad identities means that DocV can passively incorporate risk signals to correlate an identity to device, phone, behavioral biometrics, geolocation, and other indicators in the background.

For a deep dive on more onboarding challenges and how to solve those pain points, download our new ebook here.