Will AI Render KYC Useless in the Future?

Artificial intelligence is revolutionizing many industries, and sometimes this means that certain technologies and best practices become outdated and irrelevant. There is some talk that this is what is happening with KYC (Know Your Customer) processes, which are getting outshone by Generative AI.

GenAI refers to any kind of artificial intelligence that is able to create audio, video, text, or images. It’s an incredibly smart and progressive system, and is able to be used to generate very realistic virtual identities. These identities are hard to distinguish from real ones by KYC, and has caused many to come to the conclusion that Know Your Customer tech is no longer able to tell what is real and what is fake. If true, this obviously has big consequences.

For example, KYC processes are used by banks to protect against things like corruption, terrorist financing, fraudulent activity, money laundering, and more. In essence, it ensures that clients are who they claim to be. If GenAI is leading to KYC making errors, then this could lead to huge financial losses for banks and their clients, and create the need for new forms of verification and protection. Some sectors have already moved away from KYC, instead allowing customers anonymity. One such example is no-KYC casinos, which are often based on digital currencies and provide secure and private gambling for users.

A key part of what makes Know Your Customer work is the biometrics, such as having to show facial features or a fingerprint in order to gain access to a private account or information. At one point, biometrics was foolproof, but these days there are even ways to spoof this technology. Generative AI is able to be used to create “deepfakes”, the purpose of which is mostly light-hearted and humorous, rather than for anything insidious. However, they have the ability to verify an account deceitfully.

KYC processes work in a fairly simple way: either an individual will pass or fail, depending on how much an original item matches up with what is presented. For example, if an AI-replicated voice is close enough to the one initially provided, which is easy enough for Generative AI technology to do, hacking is more than possible. There is simply a lot of margin for error, which is what leading experts to suggest that KYC will soon be rendered useless by AI.

However, this does not mean that KYC is redundant. As of 2024, GenAI is not able to replicate fingerprints or retina scans, so banks or accounting firms that require this should be able to continue providing security for their clients and themselves. In addition to requiring these scans, asking for personal financial information to gain access is also an additional step that could be put in place. Unfortunately, all signs point towards a zero-trust approach being necessary if certain sectors want to remain reputable.

This article does not necessarily reflect the opinions of the editors or management of EconoTimes.