How the finance industry should be preparing for the generative AI revolution

The finance industry is more reliant than ever on efficient technological performance. Recent years witnessed a goldrush to digital systems and platforms as finance businesses sought to modernise and outperform the competition. The Market Research Reports Service predicts that the global digital banking market will be worth USD 19.2 billion by 2028, a massive rise from its USD 8.4 billion valuation in 2011.

Now, new advancements in Artificial Intelligence (AI) means the financial industry’s digital migration is poised to go into hyperdrive. Although a relatively newcomer on the block, generative AI technology has leapt ahead in a tiny fraction of the time that it took today’s social media giants to accomplish the same. Case in point: ChatGPT crossed one million users just five days after its launch in November 2023, and gained 100 million active users by January 2023 – a 9,900% increase in 60 days, crowning it the fastest-growing platform in history. To put things into perspective, it took TikTok about nine months after its global launch to reach 100 million users and Instagram 2.5 years.

The impact of generative AI will leave no industry unscathed. Goldman Sachs released a report last month predicting the AI’s impact on the global market to exceed USD 7 trillion over a 10-year period. For the finance industry, the potential here too is endless – and not just on internal matters such as task automation, better fraud detection, and more personalised services. The customer experience stands to greatly benefit too, with ChatGPT-style tools that are able to intelligently and instantly understand and answer any question without the need to engage a bank employee - for example, “Can I increase my monthly mortgage repayments by £100?” or “What is the best investment plan that I can safely afford?”. These revolutionary capabilities are transforming operational systems and processes from the ground up, slashing resource needs, and opening up endless new possibilities that we are only just waking up to. There is little doubt we will see a paradigm shift in job functions over the next several years, as well as entirely new ones created.

But for its true potential to be realised – and effective – it is important that generative AI has access to high quality data. Many companies still struggle with poor data quality due to reasons ranging from lack of team resources to siloed data, often stored in multiple locations. In fact, research shows that improving access to data distributed across various platforms has become a priority for more than 60% of financial leaders. What’s more, financial institutions face tightening regulatory legislation all over the world, heightening the need for these companies to coral their expanding data sets.

This is why finance businesses need to seriously consider a data fabric architecture before they pile onto the generative AI bandwagon. The reason for this is simple. Data fabric is the most powerful data management solution today, leveraging AI to handle data at scale and in real time. 

Data fabric works by using machine learning (ML) algorithms to automatically retrieve data from multiple sources. This data is then integrated within a centralised framework known as a virtualisation layer. From there, knowledge graphs and ontologies can be created semi-automatically, turning otherwise technical, ambiguous data that may be very difficult for humans to process, into data that is unified, linked, readable, accurate, of high quality, and giving it business meaning. And because it provides a single environment for accessing and collecting all a company’s data no matter where it’s located or stored, no data goes to waste. This means you are able to access fully trusted, fully accurate data, and in turn trust the AI model built with it and its insights.

A bank developing new business strategies, for example, might want to examine customer data including demographics, transaction histories, and account balances. By connecting all these data points together in an automatic way, the bank easily gains a comprehensive view of its customers, thereby enabling more informed and quicker (and accurate) decision making.

On a larger scale, ontologies are important for semantic interoperability – providing a common vocabulary to describe and exchange data between banks, investment firms, regulatory agencies, and the like. Standardised, linked ontologies allow systems to exchange and federate data with universal meaning, helping to future proof the data and making human-to-machine and machine-to-machine communication much more fluid. 

For the financial industry, which handles more sensitive and confidential information than most, accuracy is king. Poor data quality and bad data governance are often the reasons for issues ranging from inefficiencies to bad decision-making. Having a comprehensive view of all data also enables easier data governance, which is crucial when it comes to legislative and regulatory compliance. When reporting transactions in shares, derivatives, and bonds, for instance, an investment firm can rely on a data fabric to seamlessly unify all the relevant data into the accessible business data layer, expediting an otherwise tedious retrieval process. Moreover, GDPR stipulations such as the masking of Personally Identifiable Information can be accommodated using a data fabric, which automatically applies governance policies.

Right across the spectrum, the finance industry is waking up to the benefits of generative AI. But for companies to truly harness its power, they need to first realise that the technology’s prowess is contingent on high quality, accurate data. There is no question that data fabric is the best tool out there to help finance businesses fully embrace the next stage in its digital revolution.