Digipass® Smart Sense

The role of risk signals is evolving

Historically, risk signals have been used to compensate for the inherent weaknesses of passwords. However, they have proven ineffective against attacks such as adversary-in-the-middle (AITM) and phishing.

With the availability of FIDO (Fast Identity Online) passkeys and security keys, organizations now have access to convenient and robust protection against these threats.

However, new risk signals are needed to address different scenarios, such as:

1. 1. Reducing the need for step-up authentication with new devices.
2. 2. Prompting users to create passkeys only when there is strong evidence that the legitimate user has signed in.
3. 3. Protecting users against advanced and emerging scams.

As an add-on to the Digipass S3 Authentication Software, Digipass Smart Sense leverages AI and machine learning technologies to analyze typical user behavior and generate an anomaly score that can be utilized in the Digipass Rules Engine. This enables organizations to tailor the use of passkeys based on the anomaly score associated with each authentication event.

Adaptive use of passkeys

For example, when the anomaly score is low, indicating that a legitimate user has been authenticated, you can prompt the user to create a passkey through Digipass Registration Rules.

The anomaly score also allows you to set User Verification to “required” in cases where suspicious activity is detected, while maintaining a default setting of “preferred” when no anomalies are present.

This approach strikes the right balance between security and user convenience, particularly for users operating laptops in clamshell mode, where accessing the fingerprint sensor may be difficult.

Additionally, the anomaly score can help determine whether to trigger step-up authentication when a synced passkey is used on a new device for the first time, or when the passkey provider does not supply an authentication intent signal (refer to NIST SP 800-63).