Facial Recognition: Banking and Fraud, Face to Face

One of the most noteworthy advancements that artificial intelligence has brought forth in recent times is the ability to identify an individual based on an image of their face (it can even be from a specific frame of a video). This technique, commonly referred to as «facial recognition,» has numerous applications, particularly within the banking sector. It has been employed to mitigate fraud volume, a growth that has occurred recently, driven precisely by the acceleration of digital life.

Indeed, since 2020, banking users have massively shifted – out of necessity due to pandemic-related restrictions – towards digital channels. This shift has resulted in an unprecedented surge in cyberattacks involving password theft through social engineering tactics to pilfer account data, and even fraudulent transactions aimed at stealing money before the bank can detect the anomaly. According to data from the Cybercrime Specialized Unit (UFECI), in Argentina alone, fraud and scams related to bank accounts – such as theft of home banking access credentials or credit card information – witnessed one of the most substantial increases last year: a rise of nearly 3,000%.

How Facial Recognition Works

Facial recognition relies on unique biometric elements associated with a person’s face and expression to identify them uniquely. Due to the mathematical patterns it uses to authenticate the user, it is considered one of the most secure and effective mechanisms available to date. Machine learning and deep learning are of particular importance within AI, as they allow the algorithm to learn from its mistakes and continuously improve accuracy when identifying a face or verifying a user’s identity.

 Virtual and Physical: Applications of Facial Recognition

Banks are already embracing facial recognition across various use cases. Firstly, it grants access to mobile applications – often serving as a secondary authentication factor, an additional security layer beyond traditional elements like usernames, passwords, or PINs. The same applies when logging into virtual platforms or home banking.

In more advanced scenarios, it can be employed for purchasing products and services. For instance, by comparing a selfie with document images, customers can have their accounts authorized in less than five minutes, without even visiting a physical branch.

However, the benefits of utilizing this technology extend beyond these instances. It can also be integrated into ATMs – the system can identify whether someone is legitimate in under a second – and even for controlling employee access to restricted areas.

Card Not Present Fraud

One of the most prevalent fraud models is card-not-present transactions: when a customer conducts an online transaction involving a payment method. A recent global report, with a specific focus on this issue, estimated that the percentage of potentially fraudulent operations in this mode is notably high: 11% of total transactions in the first quarter of 2021.

Can Facial Recognition also address this issue? The answer is affirmative: if an online commerce platform incorporates this tool into their payment strategy, through a pre-confirmation approval solution, the majority of irregular or suspicious transactions could be halted before generating charges and chargebacks – both of which are significantly costly for the financial system and the involved merchants.

 The Benefits of Authentic Identity

Implementing facial recognition by banking institutions yields benefits on various fronts. For customers, it reduces risks and enhances peace of mind when conducting operations online, particularly during these transitional times where a significant portion of the population is still adapting to and learning to utilize new models.

From an institutional perspective, the reduction of costs related to fines, chargebacks, and administrative management of the vast volume of fraud is not the only advantage. It also delivers an improved overall user experience and prevents reputational issues that arise when scams are executed on a larger scale by exploiting the vulnerabilities of specific banks.

In this way, facial recognition technology «reveals itself» as a solution to the challenge of fraud within the realm of banking and financial service