From unknown power to trusted tool: how will AI evolve in the financial services industry?

While it may seem to some that AI has suddenly burst onto the scene, it was born in the 20th century and has been evolving ever since. The release of ChatGPT and Bing’s chatbot this year marked a significant growth in AI’s power, particularly for Large Language Models (LLMs), which learn faster - with less data and training required.

In previous years, AI has been clunky, slower, and required a large commitment of resources and training. On top of all this, it could be unpredictable. In 2016, Tay, Microsoft’s AI chatbot, backfired within hours of launch, with Tay citing extremist and racist views.

Unpredictability, however, is not something the financial services industry can risk. Stability and easily evidenced outcomes are expected when consumers entrust platforms with their money. The Financial Conduct Authority (FCA), the regulator of the industry, has noted while innovation is encouraged, it must be done safely and responsibly.

As AI matures, our industry can look to leverage its power with more trust. The burning question, however, is how? Let’s look closer at a few notable predictions.

Using data and insights to uncover the why:

Data analysis and number crunching are nothing new for AI. However, as the speed AI can process increases and the cost to do so decreases, we’ll be able to process large, structured data sets with less time spent on training. A financial analyst, for example, could use AI to scan financial statements in bulk for differences - analysts can then spend time investigating any flagged anomalies rather than scanning documents, allowing the team to focus resources on extracting value.

Similarly, it’s expected that more advanced models of AI will be able to provide impactful insights on financial patterns. Rather than simply flagging at what time of the month you fund your account, if AI can identify that your deposits between certain dates, outperformed the rest of the year over a significant period, consumers can then use this insight to compare activity, learn and improve their decisions. Without AI, they may never have realised one technique outperformed the other. In this case, uncovering the why is when we can unlock the power of AI.

Summary and extracting data:

Linking back to the likes of Large Language Models (LLMs) and ChatGPT, for financial services to use it with confidence, we have to be sure it can decipher tone and intent across all mediums: text, video, images etc. - something humans themselves still struggle with. If it reaches this stage, AI can be used to simplify and condense large financial reports and jargon, to make information more accessible to consumers at a much lower cost than if an organisation hired tailored teams to do this.

It could go beyond this too, analysing tone and intent in communications, helping compliance teams to ensure communication with customers is fair and compliant.

Supporting customer service

While AI is currently used in customer service departments to field inbound requests and direct customers to the correct department, as its capabilities evolve, we’ll likely see its role in communications continue to increase. A human mimic on the end of the phone might be disconcerting at first, but it could help you to find the exact information you need at a quicker rate than a human.

Tailoring behavioural analysis:

Email campaigns are an everyday example of behaviour analysis, where organisations analyse when consumers are most likely to open their mail. However, to suggest an investment provider could detect what time of day you should make an investment decision sounds almost make-believe. Push notifications and prompts are already encouraging you to engage at a time they estimate the general population is going to be available - being able to do this on a tailored basis will be the next step, and we’re heading in that direction quicker than we think.

Financial Advice: 

Finance is so much more than cash in your wallet. How you spend and invest is personal. Entrusting someone for advice is deeply personal and intimate. Considering this, I don’t think the question is whether AI will - at some stage- be able to offer financial advice, it’s whether humans would use it.

The future of AI in finance 

AI is going to change how financial services work. How quickly it will, and in which areas, is the unknown for now. Short term I expect its skills in analysis and processing tasks will be adopted by many. While its role in communications and customer service will likely take many more years as more advanced models emerge. For now, the financial services industry needs to be sure we’re leveraging a tool we understand, not an unknown power.