「金の借り手にも貸し手にもなるな」： AWS London FSI Analyst Day 2024を振り返って

While the works of William Shakespeare may not paint the financial services industry in a very positive light, banking, payments, and insurance were nevertheless centre stage at the Globe Theatre in London for the latest AWS FSI Analyst Day. Very much built on the idea of “show me, don’t tell me”, AWS brought together a collection of clients, partners, and its own experts to share how working with the vendor and its partners was bringing transformational change to their organisations.

Taking the lessons of the day’s speakers, there are three key conclusions.

Technology is not the brake on product and process innovation it once was

One of the most interesting viewpoints shared in the sessions is a reversal of the old view that technology limitations were holding back innovation in the industry. Indeed, there is a clear view that the potential of cloud technologies, machine learning, and other forms of AI to support innovation is now outpacing the appetite of business and risk teams to implement it. In other cases, the ability of regulators to keep pace with changing technology is also a barrier to progress.

One example of this was the case of DISCAI, the AML fintech founded by KBC Bank. The approach it takes to monitoring AML risk in the bank is so far ahead of some of the bank’s peers that the regulator requires it to provide reporting based on the scenario-driven model that it has sought to move away from.

Nationwide Building Society pointed to a similar dynamic with its lines of business. The organisation has a clear strategy around using cloud technology to transform the bank, and reported strong progress across several areas. When it came to product enhancement, Nationwide noted that the situation around innovation is no longer a case of the business asking what outcomes the technology can support, and much more about deciding what enhancements can be deployed successfully to customers.

Many are investing today to build the foundations for data-led innovation tomorrow

Another important theme was the importance of investing in the quality and utility of data to support transformation efforts in the future. The principal near term goal of the Blueprint 2 project at Lloyds of London is to have granular, standardised, and centralised data on the policies and claims written across the 100+ syndicates using the exchange. In some cases, this involves digitising data contained in pdf documents going back 20-30 years, as well as a broader change management program to encourage adoption of a new digital platform.

This work is the key enabling pillar to delivering a range of machine learning and GenAI supported services in the future. However, and as was noted on stage, “you can’t have GenAI without digital information”. This is a situation in which many Celent clients find themselves, which is driving similar initiatives and investments across the banking industry.

While GenAI may get the hype, there are huge gains to be had from investing in other forms of AI

A discussion about Gen AI is an essential feature of any conference or analyst event these days. Given the potential of the technology, this is clearly justified, but the sessions highlighted two important perspectives about where the industry is in its adoption of the technology.

The first is the degree of caution around deploying GenAI in directly customer-facing use cases. This doesn’t mean that there aren’t many areas in which it can (and will) bring value; NASDAQ spoke about using the technology to support anomaly detection, and as a natural language interface for running queries across large datasets for example. In addition, Nationwide also highlighted the potential to support agents in the contact centre or in creating templates for responding to some customer communications. Bunq bucked this trend slightly by showcasing a GenAI-powered customer-facing chatbot, but confirmed that there are strict limits on the scenarios the bot is able to address.

However, looking at the data-led use cases shared in the sessions that are really making an impact right now, these are being supported by what’s best described as ‘classic’ AI. To pick out just two examples, more traditional machine learning technologies are at the core of KBC’s approach to transforming its AML monitoring, while Databricks referenced the growing demand for using AI for hyper-personalisation among its financial services clients.

The key lessons from the day are clear. We’re only at the relative beginnings of the data-led transformation ahead, and the most forward-thinking institutions are investing in laying the foundations for this future change. In addition, while GenAI will support a lot of the innovation ahead, the AI story will be told by a far larger cast of players.