Generative AI will not be a disruptor for financial services – but it can be its enabler

By Jared Johnson, Strategy Director at global digital transformation consultancy Kin + Carta

Whatever the industry, Gen AI is having an impact – for better or worse. And like with all disruption of this nature, questions are being asked. The financial services sector is no different and is keen to find out if it will lead to the next wave of disruptor companies like Monzo and Starling.

However, finance is not retail, where the technology is making huge waves. Financial organizations are highly regulated and require 100% accuracy when it comes to transactions and recommendations. In short, they are based on trust, and trust is currently in short supply when it comes to AI.

That is not to say there is no place for Gen AI. It will just be internal rather than customer-facing for the moment. A customer service agent might use it to recommend questions to ask the customer or products that might be suitable, but currently there still needs to be a human in the loop to ensure the insights being offered are suitably trustworthy.

So, while it will no doubt prove to be an invaluable tool, it will mainly remain in the background, working to further enable employees within the financial sector to focus on fine-tuning and improving internal systems. 

There may be some customer-facing use cases, but these will be where there is little or no risk. A bank’s marketing department could use generative AI to create new written content and generate photo-realistic images for advertisements in seconds rather than days, for example. These customer-facing use cases are low risk because a human can be in the loop to tweak text or re-generate an image until it’s just right. 

However, giving the customer direct access to Gen AI is far riskier. For example, giving a Gen AI-derived recommendation to a customer about what is best to do with their money requires inputs from multiple parts of the business (risk, compliance, CRM, customer data, business rules, etc.) which means that the Gen AI’s recommendation can only be trusted if the proper processes and data foundation exist in the organization. 

Gen AI needs the right inputs from your business to interpolate a trustworthy and compliant recommendation to a customer. GenAI doesn’t solve these foundational problems, like lacking a data foundation or clear business rules,  but rather is an enabler on top of your foundation

Disruption incoming? Probably not

For banking and other financial services, technology is often not the primary challenge. This is another reason why they are often slower to adopt emerging technologies.

Instead, the challenges are about building a brand, creating trust, and having the infrastructure in place to deliver an excellent customer experience. They need to put all these pieces together. And generative AI is most successful and scalable when it is based on trustworthy data, which existing organizations possess to varying degrees and new disruptors would struggle to acquire quickly enough to be of value.

Those financial businesses have the data, and the existing level of trust, so for them AI could be an enabler to leverage it more effectively.

Gen AI’s strength is its ability to free up the financial services sector of menial tasks such as data admin, freeing people to focus on broader strategies of customer experience. But even then, it will still have to leave humans in the driver’s seat – which is crucial when dealing with customer financial data.

It is unlikely to create a new wave of disruptors in the same way as online and mobile banking did, simply because people – quite rightly – do not trust it yet to stand on its own.

Building better financial services

Given that the value of generative AI lies in its flexibility in the internal, employee-facing arena, that is where most financial services organizations should be looking to build solutions. Where there is a cumbersome process, or where creating a range of hundreds of different options to choose from is useful, AI can shine.

But don’t rush to assume it can be part of the customer-facing process just yet – and don’t assume that it will rewrite and redefine the industry. At least for the immediate future, the technology is best employed for improving what exists already.


Explore more