How can I recognize signs of synthetic identity fraud before activating an account?

A KYC verification needs to treat synthetic identity fraud This presents a different problem from simple document forgery. In this type of fraud, the registration is assembled with a combination of real and invented information to fabricate an identity that appears coherent when each piece of data is examined in isolation.

Therefore, the defense should not be limited to asking whether the CPF (Brazilian taxpayer ID), phone number, document, or face "passed" a check. The goal is to prove that the elements form a consistent identity and that this identity truly belongs to the person trying to activate the account.

Summary

• Synthetic identities can combine real and fictitious data into a seemingly legitimate registry.
• Good proof of identity separates resolution, validation of evidence, and verification of the link to the person.
• CPF (Brazilian taxpayer ID) and phone number, biometrics, registration history, SIM swap, device, and anomalies work best together.
• The process should reduce fraud without turning every legitimate customer into a suspect.
• The operation needs to track fraud prevention, false positives, approvals, abandonment, and manual review.

How does synthetic identity fraud get past seemingly normal registrations?

The most common mistake is to imagine that a sophisticated fraud always involves an obviously falsified document or absurd data. That's not always the case. Federal Reserve explains Synthetic identities can combine a legitimate identifier with fictitious names, addresses, or dates of birth, and precisely because of this combination, they can evade traditional verification processes.

For a digital company, this changes the central question. Instead of "does this CPF [Brazilian individual taxpayer registration number] exist?", the operation needs to ask: "does this set of information represent a real, unique, and coherent person?"

When we founded ZapSign, two principles were very clear to me: the experience needed to be simple, and the solution needed to offer genuine legal security. I see fraud prevention by the same standard. Adding more steps to onboarding might give a sense of security, but poorly calibrated security translates into cost, abandonment, and support issues. The best architecture is one that increases trust in identity without forcing all users to go through the most arduous process.

This reasoning also helps to separate Electronic identity and signatureA signature records a statement linked to a process; proof of identity seeks to demonstrate who is on the other side. The greater the risk of the operation, the more evidence makes sense. The page itself... signatory's identity It shows how different methods can be combined according to the level of risk, instead of applying the same authentication to any document or journey.

How do I structure the proof of identity process before activating my account?

A consistent journey can be organized into four decisions: resolving who the person claims to be, validating the presented attributes, verifying if the evidence belongs to that person, and applying additional risk signals. The value of this sequence lies in preventing a single “match” from being treated as automatic approval. A valid CPF (Brazilian tax identification number) confirms one piece of data. A compatible selfie confirms another link. An old and consistent phone number adds context. No single piece should bear the responsibility for the entire decision.

1. Resolve the claimed identity.

In the first stage, the company needs to gather the minimum amount of information necessary to distinguish that identity within its own user population. Name, CPF (Brazilian tax identification number), date of birth, and contact information can form this basis, but collecting more data does not automatically mean knowing the customer better. combination of features It must adhere to the actual risk of the journey. Excessive data increases exposure and friction without necessarily improving decision-making.

2. Validate attributes and evidence

Next, the question is whether the data and documents presented are authentic and consistent with each other. This involves background checks, document examination, reliable databases, and consistency between fields. GAO points out Synthetic identity fraud is a growing concern in digitized financial environments, and it describes electronic verification services that compare name, identification number, and date of birth. The operational lesson is simple: a single piece of data can be correct and yet belong to the wrong story.

3. Verify the link between the person and the evidence.

It is at this stage that biometrics, proof of life, and facial comparison cease to be "extra layers" and begin to answer a specific question: does the person present verify and match the evidence used in the registration? ZapSign already brings together resources of this type in its... evolution of authenticationsThe gain lies not in accumulating obstacles, but in transforming independent evidence into a more reliable conclusion about who is carrying out the action.

Which signals should be combined against synthetic identities?

Synthetic identity fraud requires layered controls because the attacker is precisely trying to produce enough consistency to bypass superficial checks. GOV.UK recommends To verify the existence of an identity over time, to compare suspected identities with authoritative anti-fraud sources, and to confirm that the identity belongs to the person claiming it. This leads to a design in which history, phone, biometrics, and device complement each other.

Signal	What to watch out for	Use in decision-making
CPF and phone number	Compatibility between ownership, registration, and contact information provided.	Detecting unlikely combinations or borrowed data
Biometrics and liveness	Facial recognition and real-life presence throughout the journey.	Confirm the link between the person and the evidence.
History or tenure	Existence and activity of identity over time.	Identify newly manufactured airfoils or airfoils without a consistent trajectory.
YES swap	Recent SIM card replacement or unusual line change.	Increase the risk before trusting possession of the phone.
Device fingerprinting	Device, browser, recurrence, and association with other registrations.	Finding networks of accounts and repeated patterns
Anomalies	Speed, sequence, timing, and unusual combination of actions.	Forward exceptions for step-up or review.

When we demonstrated ZapSign's use of Open Gateway, the proposal was precisely to use the existing telephone network infrastructure to validate signer information and combat identity fraud without creating new friction for the user. I like this example because it exposes a product choice that should guide every anti-fraud project: whenever possible, improve signal quality behind the scenes before asking the customer for another action.

This logic also applies to integrations. A company that already performs onboarding, CRM, subscription, and risk analysis in different systems needs to make the signals communicate. electronic signature API and integrations between tools They help reduce operational islands. For more sensitive flows, a background check It can be included as an additional step, provided there is a legitimate purpose, proportionality, and a clear reason for that collection.

How can we reduce fraud without increasing false positives?

Blocking more doesn't mean better protection. If the policy rejects legitimate customers, increases manual review, and transforms minor discrepancies into automatic rejection, the anti-fraud system ends up creating another type of loss. The best design uses risk levels. A consistent registration follows with low friction. An intermediate case receives additional verification. A set of incompatible signals goes for review or blocking. This model allows for calibrating security without treating every exception as completed fraud.

In operational practice, I would monitor at least five KPIs together: fraud prevention rate, false positives, approval rate, abandonment rate, and percentage of manual review. Focusing solely on fraud prevention creates an incentive to tighten the funnel until conversion is destroyed. Focusing solely on approval does the opposite. Balance is achieved when security and experience are on the same dashboard, and each rule change can be compared to its effect on the business.

• Fraud prevented: It measures the ability to prevent unwanted activations.
• False positives: This shows how many legitimate users were treated as high risk.
• Approval: It tracks the percentage of journeys successfully completed.
• Abandonment: It helps to identify friction points that drive away potential customers.
• Manual review: It reveals how much human cost politics is shifting towards operations.

Check out these related articles as well:

• Understand how identity validation works in digital journeys.
• See how facial biometrics can support authentication and fraud prevention.
• Learn about the main components of an anti-fraud system.

A good defense begins before the account is activated.

The best time to discover that a registration doesn't match is before granting access, limits, signatures, or transactional capabilities. This requires proof of identity that combines resolution, validation, linking, and contextual signals, with controls proportionate to the risk. Companies that already operate digital processes can connect this logic to... signatory validation and to the subscription processes, preventing security from only appearing after the incident.

In the end, fight synthetic identity fraud It's not about turning onboarding into an interrogation. It's about building enough trust to separate a legitimate person from a fabricated identity, using the right evidence at the right time. If your operation needs to reinforce this point before activating accounts or formalizing documents, learn more about... ZapSign ID and assess which layers make sense for the actual risk of your flow.

Frequently Asked Questions (FAQ)