As an identity verification company, Persona is constantly thinking about fraud — how to detect it, how to fight it, and how to prevent it for our thousands of customers like Square and Travelex. (If you need proof, just search our blog for “fraud" and see all of our posts!)
But why? Because the vast majority of fraud, no matter what form it takes or the role AI plays, is ultimately conducted by a human. By taking an identity-first approach to fraud, you can attack the root of the problem, rather than the symptoms — think the difference between blocking a fraudster at onboarding versus catching chargeback fraud on your platform after the damage is done.
Against a backdrop of freely available AI tools, this relationship between identity verification and fraud mitigation becomes more important than ever — not just for us or our customers, but for all companies facing sophisticated fraud that looks different week to week. In fact, in our upcoming fraud survey, the majority of respondents across all verticals cited identity fraud as a top priority.
Breaking down this burning problem, we’re seeing three interwoven problems that companies face today.
Problem #1: Fraud is more adaptive.
Gone are the days of fraudsters just using doctored IDs to bypass verification checks. Instead, they’re now using generative AI tools to create synthetic IDs and deepfakes to evade detection. These tools enable fraudsters to both do more and evolve faster, and as a result, synthetic fraud losses are expected to hit nearly $5B by 2024.
Rather than relying solely on image-based techniques, companies need to leverage every signal they have at their disposal to determine if someone is a fraudster or not, especially since bad actors are constantly finding new ways to get around verifications. This game of cat and mouse means you not only need to be able to add different types of verifications and checks, but you also need to be able to do so quickly.
The problem we’ve heard is that teams want to incorporate more risk signals in their KYC process but are locked into rigid platforms that only specialize in one type of verification, like documentary verifications. As a result, they can't keep up with evolving fraudsters or are forced to go between multiple systems to truly understand the riskiness of a user.
Problem #2: Fraud is more pervasive.
Our customers, especially those in fintech, have always understood the importance of robust KYC during onboarding to prevent fraudsters from taking advantage of new account promotions or committing account creation fraud.
But across all verticals, bad actors are increasingly striking at key moments after onboarding. These sensitive touchpoints vary by company, whether it’s at withdrawal for a fintech, payout for a marketplace, or upon publishing a new listing at a property tech company, but the vector of fraud — account takeover or activation of a network — is generally the same.
Fraud teams often have no systematic way to monitor these accounts or touchpoints. As a result, they often find themselves reacting to fraud fires or implementing stringent verifications that deter bad actors and good users alike.
Problem #3: Fraud is more costly.
Not only is fraud costly in the obvious ways, but to fight it, many organizations now spend a lot of effort coordinating different tools or maintaining a standalone orchestration tool instead of throwing people at the problem. Our survey supports this, finding that teams prefer reconfiguring or adding new fraud prevention technology rather than hiring more staff.
Still, relying on tools takes a toll: analysts tell us they either have to manually toggle between teams and systems to respond to incidents or managers will leverage orchestrators that require intense engineering resources whenever a new signal or process needs to be introduced. In a time of “do more with less,” fraud teams don’t have the resources they need to adopt the strategies they want.
Catch more fraud with less overhead with Persona
While there’s no silver bullet to these challenges, we believe in the value of one unified identity platform that lets you proactively combat fraud across all of these dimensions. We’ve spent the past year intentionally focused on helping our customers better fight fraud, culminating in our end-to-end approach to fraud prevention.
Stay ahead of shifting fraudsters with more risk signals from Persona and Marketplace partners
While we’re constantly improving and expanding our own library of verification options and reports, we know that agile fraud teams need to be able to incorporate risk signals from all of their vendors to assess user riskiness holistically. Better yet, they need to be able to do so quickly.
With the 20+ integrations in our Marketplace, you can seamlessly leverage the expertise of top data platforms like SentiLink, Telesign, and MX in specific domains like synthetic ID, phone risk, and bank account verification without having to build your own integrations or leave the Persona platform. By pairing our in-house risk signals with those of our partners, you can better identify and keep pace with third-party fraud, no matter what form it takes.
Dozens of our customers use Marketplace integrations, either directly within their KYC flow or as an additional risk signal in Workflows, to catch more fraud without adding more end-user friction. For example, a fintech focused on instant payments flags more bad actors by integrating SentiLink and Persona.
After a user has gone through KYC, that fintech automatically calls for SentiLink’s synthetic ID analysis, which is returned in milliseconds. Depending on the results, which include a composite score for falsified and manipulated identities (also known as synthetic fraud), they can automatically make a decision or kick off additional reviews, ensuring its team’s resources are allocated most effectively.
Catch bad actors at every stage of the customer life cycle without unnecessary friction
To help organizations fight fraud from onboarding and beyond, we offer active verifications or silent monitoring at any point of the customer journey. The end result? You can continuously confirm users while catching more bad actors — no matter where they are in their user life cycle.
Use Dynamic Flow to add verification friction at risky touchpoints (e.g. onboarding and reverification) based on real-time risk signals instead of blindly verifying all users the same way. You can layer different verifications based on real-time risk signals that span Persona and third parties instead of relying on rigid risk segmentation. For example, if a user has a different device fingerprint than what they had at onboarding, you can ask for an ID and check it against the original one submitted at onboarding to ensure their account isn’t being taken over.
You can pair this with Graph, which can screen and identify networks of bad actors before they strike for passive deterrence throughout the customer life cycle. With link analysis, you can monitor accounts to quickly connect and surface clusters of risky behavior.
For example, if someone tried to create an account for Neighbor to score a referral bonus, then create 50 more accounts, Graph could surface all the accounts associated with the original account’s IP address and allow Neighbor to block all the accounts at once — along with future linked accounts.
Respond faster to fraud attacks with no-code orchestration for all of your tools
Lastly, you can now put your fraud team in control with a multi-layered fraud approach that can be configured inside Persona without code.
With access to a suite of fraud detection, prevention, and investigation tools that can all be orchestrated by Persona, teams only need to learn one tool and go to one place to build the processes they need.
But breadth doesn’t come at the expense of depth for our fraud prevention capabilities. You can also use Workflows to build any type of fraud prevention process, whether it’s complex rulesets across multiple vendors or a bespoke monitoring process.
In one case, we partnered with Sonder to build ML-powered risk detection during its bookings process. Its trust and safety team was able to easily bring in and add additional signals to power this model — from guest data collected by Persona to booking information like length of stay and how far a guest books in advance. As a result, Sonder was able to respond faster to rapidly changing fraud observed during its bookings process, with the added benefit of automating 70% of manual reviews.
Whatever form your fraud prevention processes take, Persona can power it, from verifications and checkpoints along your customer journey to monitoring and investigations — with orchestration to unite it all together.
It bears repeating: fraud is inevitable. But by layering together the capabilities of our products, customers like Neighbor, Sonder, and Branch move faster to catch and block the fraudsters they’re seeing today while being empowered to prevent the fraud of tomorrow.
If you’re interested in supercharging your fraud-fighting abilities, reach out to your CSM or get a demo today.