Katharine Wooller is Chief Strategist – Financial Services, Softcat plc, a FTSE-listed IT company.
Discover top fintech news and events!
Subscribe to FinTech Weekly's newsletter
Read by executives at JP Morgan, Coinbase, Blackrock, Klarna and more
Few topics are as polarising as AI; verdicts range from, at the more positive end, the next frontier of human progress, a technology solution looking for problems to fix, or, at worse, the potential to create the end of mankind.
As a Chief Strategist for Softcat, who supports 2,500 financial services firms through IT services and infrastructure, I have a privileged front-row seat in watching innovation unfold across the whole spectrum FS&I firms.
First out the gates, there has been strong uptake in quant hedge funds, who embrace the significant investment in AI for improved returns, and also insurance, which benefits from huge amounts of data – both can easily justify clear uses cases with a strong ROI.
Financial services firms have been doing mathematical modelling and machine learning nearly a decade before AI was marketed in its current guise, but recently the shear performance of AI infrastructure has stoked a strong uptake by quantitative trading funds and insurance and wealth management firms, all seeking benefit from the huge amount of data now available to them.
Moreover, a lot of what is sold as AI is simply the next incarnation of automation.
Whilst we see huge interest in AI across all types of financial services firms, based on the huge potential of the technology, we are ultimately at the foothills of adoption. Further there are hugely variant use cases – a tier one bank will deploy AI very differently to, say, a ten-branch localised building society.
I often see differing appetites within the same organisation, with boards, the younger more digitally savvy generations, and operations/finance functions often more welcoming to the idea, than, say, compliance colleagues. Concerns raised often include the “black box” nature of the technology, worries around ethical deployment of AI, and lack of regulatory clarity.
There are, however, clear patterns emerging in what makes for early uptake and strong levels of usage. Successful firms have a strong strategy for adopting AI, setting up centres of excellence and making sure their data is in an appropriate state from the get-go; these sound like small undertakings, but they are the bedrock of successful innovation.
We often see the first use case to deploy in productivity tools such as ChatGPT, Co-pilot, or Claude, which are often the entry point for many colleagues in embracing the idea of AI, and sometimes dryly referred to as the “gateway drug”!
Culturally, adopting AI can be a huge departure from the status quo, and highly effective leadership teams will be looking to future proof their organisations. A forward-thinking HR strategy is paramount, building internal AI capabilities and expertise, focusing on applicable skills, expertise and encouraging knowledge sharing. A long-term view will need to be taken on redeploying colleagues whose roles are displaced by AI driven efficiencies.
There is rightly much focus on the AI value add; there are some banks who have hundreds of potential use cases and navigating which to get into proof of concept, and roll out more widely, can be challenging. Best practise, for such a new technology, is only just emerging. In the first instance, shifting through a huge number of potential use cases to prioritise those which offer the greatest value creation can be overwhelming, and ruthless triage can be made based on impact, cost, feasibility, and alignment with broader business objectives, to evaluate potential ROI.
There needs to be a well thought out measurement framework to evaluate AI projects, with relevant KPIs, robust data collection methodologies, and clearly defined reporting mechanisms. Once an AI project is part of BAU, there needs to be a policy of continuous iterative development over time to maximise returns and ensure alignment with strategic priorities - again this is often a cultural feature of high performing teams.
Recently, I was invited to talk about AI with a regulator. During an industry round table, a brilliantly perplexing question was presented: “What one problem does AI solve better than anything else?” Unsurprisingly, each organization had a completely different answer, and I expect firms to be grappling with this question for years to come.
Those who cannot be strategic about AI, and deploy in an appropriate and timely fashion, will be at a significant disadvantage.
