Even inexperienced scammers are using the new generative AI models that appear nearly every week to create convincing ID verification sessions on a large scale.
Due to this change, defence-in-depth, which involves placing multiple independent checks to ensure that nothing depends on a single point of failure, is more important than ever. We’re thrilled to present robust new security features that streamline the process for trusted users while making it even more difficult for scammers to evade or circumvent identity verification.
There is no need for extra integration because these improvements are available to all Plaid Identity Verification (IDV) customers right out of the box.
Avoiding synthetic fraud is crucial as AI-generated media becomes more and more similar to the real world. That’s why we’ve enhanced our identity verification with advanced machine learning to detect deepfake injections and synthetic camera feeds during both liveness and document checks.
However, stronger camera source integrity checks now stop altered feeds before they reach your flow, while upgraded face identification algorithms catch visual anomalies associated with AI-generated content. We’ve also added numerous levels of protection to liveness and document verification, including tougher checks for document tampering and forgery, enhanced selfie-to-ID photo matching, and dynamic liveness detection that resists spoofing attempts like printed photos or screen recordings.
These improvements contribute to faster, more accurate identification checks and better protection for your users by ensuring that the person on camera is the same person on the ID. To identify recurring fraud, identify duplicate faces. The same person or digital image is frequently used by fraud rings across several accounts, sometimes with fake documents or manipulated identification.
Plaid’s Facial Duplicate Detection function assists in preventing this by classifying faces and spotting possible duplication in document photos and selfies. With a false match rate of only 1 in 1 million, which leads the industry, you receive exceptional precision without any extra friction. Additionally, you can choose to match inside particular use cases or across your whole user base thanks to numerous setup options, which make it simpler than ever to identify fake identities, shut down mule accounts, and stop repeated fraud attempts.
A new biometric tool that determines a user’s age while they are in a live session is being introduced. An important indicator of possible identity misuse or impersonation, this strong signal helps highlight significant differences between a user’s selfie, ID document photo, and stated age.
Continuing, for our machine learning models to function consistently and fairly across a broad range of contexts and demographics without sacrificing the user experience, they are trained on a variety of datasets.
To provide the appropriate level of friction at the appropriate moment, we have added new features that provide customisation of the identity verification process depending on real-time risk. These improvements, which are powered by our Trust Index, assist you in strengthening rules for higher-risk scenarios while streamlining verification for trusted individuals.
Risk-Based Escalations: Modify the verification path in real time according to user risk. Higher-risk customers are immediately bumped up for extra inspections, while high-trust individuals can get through with less difficulty.
Selfie Re-Authentication: Verify users again by contrasting a recent selfie with one taken during a prior session. Perfect for high-risk logins, right-to-work enforcement, and sensitive use cases like account recovery, making sure the person returning is the same one who signed up.
Trust Index Risk Check: With more Editor options, you have more influence over the reasoning behind decisions. Now, even if individual checks don’t trip their limitations, you can reject a user when the total Trust Index score indicates heightened risk, in addition to when certain risk categories surpass predetermined boundaries.
BrandSpur technology and information news reports that Plaid provides businesses with a quicker, more secure method of confirming identity and preventing fraud at the front door by fusing high-fidelity signals and extensive data sources with the size of our network. Additionally, these improvements are now available to current Plaid IDV clients; you can begin taking advantage of these cutting-edge safeguards right away, and your solution will only get better.
