Welcome to the machines

Will machine learning and artificial intelligence transform financial services?

Robotic process automation (RPA), machine learning, and artificial intelligence (AI) have the potential to transform financial services. They could enable providers both to reduce costs and provide a deeper, more personalised, relationship with their customers. But critics have raised concerns about the societal and ethical impact of such technologies.

At the 2018 Innovate Finance Global Summit, journalist Andrea Catherwood used Deloitte Pixel’s Remesh platform to find out more about the Summit delegates’ view on the impact of AI and automation. The audience was optimistic about the exciting potential and business opportunity of these technologies, but also stressed the need for education “to get people up to speed”. The attendees strongly disagreed that these technologies “are all hype”.

Gurpreet Johal, Deloitte UK’s Artificial Intelligence leader, then explored what AI really means, and how it can add value to financial services firms. He described AI as “the ability for computers to process data faster; get better insights; extract concepts, meanings and learnings; and interact with humans in a natural way”. According to Gurpreet, the forces elevating the potential of AI are “the exponential growth of data, smarter algorithms and machine learning, faster processing speeds and cloud computing, and ‘big tech’ companies investing in research and development”.

He cited Deloitte research suggesting an over 50 per cent adoption rate of AI and/or data analytics among financial services firms. However, he still believes there is a lot to be learned from other industries, such as retail and media. In terms of use cases in FS, he highlighted the potential for predictive and prescriptive data analytics to anticipate future customer needs and behaviours; for example, predicting future mortgage applications six months in advance, or using spending history patterns to determine when and how to interact with customers.

Jesse McWaters, Financial Innovation Lead at the World Economic Forum, went even further, opining that AI and machine learning have the potential to fundamentally transform existing business models. He hypothesised a future in which “insurers protect against risks ever happening rather than just providing cover against those risks”. Satadru Sengupta of DataRobot made a similar point when he said that “AI is not just about predictions; it’s about changing the outcomes based on those predictions”.

Google’s Kirill Evreinov described how AI was embedded in most of his company’s products, and said that the tech giant’s cognitive technologies “have seen a lot of traction from FinTechs, for fast calculations for risk and rates, but also really exciting things such as fraud detection”.

Satadru added marketing as one of the use cases with the most potential. He believes that, along with its speed and processing power, one of the advantages AI technology has over humans is that while “we are influenced by our emotions and moods when we make decisions, AI is emotionless”. However, he noted that this raises ethical questions, noting that financial services firms using AI “have to be very careful that they are not discriminating against people based on their location, for example”.

Christine Foster, Managing Director for Innovation at the Alan Turing Institute, also discussed the ethical considerations behind AI, stating that while “technology is neutral, it’s how we use it that matters… it’s not enough to simply have an ethical framework. You need to examine the specifics behind each use case”. Jesse highlighted the importance of interpretability, or explaining to end-users why certain decisions have been made. Gurpreet believes that an integral part of ensuring customers are treated fairly is “having diversity in data scientists”.

Catch up on all things FinTech

Did you find this useful?