Impact of Deepfake Technology on FinTech Applications

Introduction

The term deepfake is derived from two words ‘deep learning’ and ‘fake’. The deepfake was introduced in the year 2017 in the Reddit community discussion when a user posted a digitally altered video clip of a celebrity (Jones & Capstone, 2020). Since its inception, deepfake videos have been created by artificial intelligence (AI) or various machine learning (ML) techniques that look authentic. Deepfake content is generally produced using two different ML techniques: autoencoder and Generative Adversarial Networks (GANs). Artificial Neural Networks (ANNs) mimic the human brain in learning and recognizing patterns with the concept that the more real image samples that are fed into a dataset, the more accurate it can be replicated as a fake sample. A conventional GANs model comprises of two neural networks, a generator, and a discriminator. The generator is a convolutional neural network (CNN) that produces synthetic images creating fake videos from the original dataset. The discriminator is simply a classifier that tries to distinguish real data from the data created by the generator and attempts to analyze and distinguish the deepfake video from the real to synthesize for authenticity. The cycle continues with the two ML algorithms, with the generator network continuing to create fake videos until the discriminator network no longer can detect and authenticate the video forgery. With the increase of deepfake technology and easy-to-use sophisticated open-source applications in the market, individuals can manipulate images and videos by superimposing someone’s face, mimicking facial expressions, and synthesizing their speech. In the start, it took days and required considerable knowledge and skills to create deepfake content, but now it is a matter of hours to create deepfake with extensively accessible free online applications (Zhao et al., 2021). It is practically impossible for a human to differentiate between genuine and forged content (Westerlund, 2019).

Additionally, cybercriminals have taken advantage of the technology to misinform and defraud businesses and individuals. As a countermeasure, FinTech was adopted, the term ‘‘FinTech’’ is a combination of ‘‘financial technology’’ and was first mentioned in the 1990s by Citicorp’s chairman John Reed (Puschmann, 2017). It contains innovative financial solutions enabled by IT and includes the incumbent financial services providers like banks and insurers. Regardless, both audio and video deepfake can manifest multiple issues within financial services. There are many cases of fraudulent onboarding (criminal posing as someone else), fraudulent payment authorizations and transfers, synthetic identity, impersonation of business leaders for insider trading tricking employees into taking nefarious actions. Such as: -