Scam Signal [Developer Preview]

Scam Signal addresses the lack of visibility in the "Authorized Push Payment" (APP) fraud cycle. APP fraud occurs when a bad actor manipulates a victim into voluntarily authorizing a real-time bank transfer under false pretenses. Unlike unauthorized fraud, the transaction is technically "legal" as the legitimate account holder performs the action, making it significantly harder to detect and recover.

Traditional fraud detection systems monitor transaction data but remain blind to the external social engineering occurring via voice channels. Scam Signal solves this by providing real-time "telephony context", allowing businesses to identify a potential fraud situation in real time, thereby preventing losses that occur when customers voluntarily—but mistakenly—authorize fraudulent transfers.

Amongst the most common use cases where Scam Signal can bring benefit, there are:

• Financial Transaction Verification: Scam Signal enables the business to query for specific and unusual telephony patterns on the customer´s associated MSISDN at the same time as this customer initiates a high-value or unusual bank transfer. The signals provided by this API while the end customer is performing the transaction can indicate a high risk of "Authorized Push Payment" fraud. The bank can then trigger a step-up authentication, a proactive warning to intercept the transfer, or require a "cool-off" period before the funds are permanently released.

• Impersonation defense on account management: Fraudsters frequently impersonate government officials to gain access to sensitive accounts or tax refunds. By integrating this signal into login and profile management flows, platforms can detect fraudulent ongoing situations that could lead to, for example, end customers making undesired critical account changes. This live context enables portals to block sensitive credential updates or redirect the user to a secure, offline verification channel to prevent account takeover.

• Marketplace transaction security: Online marketplaces and P2P platforms can use this check during the final stages of a transaction to protect buyers from advance-payment scams. If the API provides suspicious signals while the end customer is attempting to pay for goods considered within the “high risk of scam” categories, the platform can temporarily pause the payment process and introduce a context-aware intervention, such as a message or a “safety hold” of funds. This friction breaks the fraudster’s psychological momentum, providing the user with a vital moment to reassess the legitimacy of the seller, and often prompts the fraudster to abandon the attempt when they realize the immediate payout is blocked.

• Scam protection for vulnerable individuals: Scam Signal insight allows businesses to identify unusual telephony patterns that coincide with sensitive processes such as the opening of a digital banking app or an investment account. By monitoring for these high-risk overlaps, banks can provide an enhanced safety net for elderly or vulnerable customers who are primary targets for romance or investment scams. This proactive monitoring allows for early intervention, potentially stopping a scam before the victim is even persuaded to begin a transaction.

Access to Scam Signal API

The Scam Signal API specification is not publicly available. To maintain the API's effectiveness in preventing fraud, access is strictly controlled.

Prerequisites

To use Identity Insights, you must ensure your account is configured correctly; see the Getting Started guide for more information on:

• Creating your account,
• Creating a Vonage application for use with the Identity Insights API,
• The different environments available and how to configure your account to use them,
• And how to use the Dashboard Getting Started UI to use the API without writing any code.

This guide will explain how to use the Scam Signal Insight programatically using cURL.

Making an API Call

Authentication for the Identity Insights API is done via JWTs, a compact and self-contained JSON token. To generate a JWT, you can use our online generator, or alternatively use the Vonage CLI. You will need your application ID and private key to generate the JWT. Once you have your JWT, you can send a request to the API.

This example shows a cURL request for the Scam Signal insight to provide real-time telephony context associated to the give phone number:

curl -X POST https://api-eu.vonage.com/identity-insights/v1/requests  \
  -H "Authorization: Bearer $JWT" \
  -H "Content-Type: application/json" \
  -d '{
    "phone_number": "14040000000",
    "purpose": "FraudPreventionAndDetection",
    "insights": {
        "scam_signal": {}
        }
    }'

Here, the status object indicates the status of the information returned for the specified phone number:

If status.code in the response is OK, you may also see the fields described in the table below. If a field is marked as 'Yes' in the 'Mandatory' column, it will always be returned when the status is OK. If a field is marked as 'No', it may or may not be returned.

Further Reading

• Read more about the Identity Insights API in the API Reference.
• If you have any questions, you can reach out to us on the Vonage Community Slack.