Industry experts weigh in on AI’s transformative role in finance

Onome Amuge

While generative AI is often perceived as a “superpowered” technology that can perform any task, a more nuanced view is that it can also act as an orchestrator for other systems, enhancing their performance. This view has been explored by industry experts, who argue that generative AI’s capabilities are not limited to simply generating new content as it also serves as a valuable tool for transforming financial services.

At the recent AI in Financial Services Forum held  in London,United Kingdom, a panel of experts explored the role of AI in revolutionising financial services. The panel tagged “Revolutionizing Financial Services”, was moderated by Viktoria Ivan, a senior data scientist at PayTech company Ebury, and included Andrew Allright, State Street EMEA head;Tim Mason, Deutsche Bank head of innovation; Dion Kraanen, M&G director of analytics; and Ash Garner  Tomoro co-founder. The panelists discussed the role of AI in transforming financial services, including the challenges and opportunities it presents.

Generative AI has been one of the most discussed technologies in the past year, with businesses across the world exploring how it can be incorporated into their operations. Although it is still in its early stages, a recent report from Salesforce found that a majority of workers are already using or planning to use generative AI.

The panelists agreed that Generative AI has the potential to revolutionise the customer experience. This was highlighted as one of the key areas where the technology could have an impact.

A great example of the potential limitations of generative AI, and the importance of personalization, was shared by Tomoro’s Ash Garner during the panel. He noted that an Australian bank was using an algorithm to recommend products and services to customers, but the messages were not personalised and did not take into account the individual’s specific circumstances.

Garner remarked, “What we ended up doing was actually building a large language model to personalise those communications into really small segments built around how you spend your money, your personality and where you are in the world. People could read it and go, ‘that’s something that’s relevant to me’. We got about 100-150% uptake on the education content for that bank as it is actually useful for customers.”

M&G’s Dion Kraanen also discussed the importance of personalisation in the context of financial services and investments. He noted that while it may be efficient to have a robo advisor generate a portfolio based on a few checkboxes and questions, this approach lacks the depth and nuance of a real conversation.

Deutsche Bank’s Tim Mason shared his insight on the potential use cases for generative AI, highlighting the various ways in which it could be leveraged across the financial services sector. He noted that it could be used to improve the customer experience, through chatbots and other interactive tools. He also highlighted that it could be used to improve corporate client experiences, such as by automating repetitive and time-consuming tasks.

Turning the discussion to the internal benefits of generative AI, the panelists were asked about potential use cases for reducing costs and improving processes within financial institutions. Tim Mason noted that one of the most valuable applications is to leverage the technology to make sense of unstructured content. This includes documents, emails, and online chats, all of which can be difficult to process and understand using traditional methods.

Kraanen offered an interesting perspective on the next generation of generative AI applications. He suggested that while the ability to process unstructured data is valuable, the real potential of the technology lies in its ability to synthesise and analyse large amounts of data to provide unique insights.

The pitfalls of using AI in finance

One of the key risks of using generative AI is the potential for the technology to do or say something that is inappropriate or contrary to the intended purpose. This is particularly relevant in regulated industries like financial services, where there are strict rules and guidelines on how information can be presented and communicated. Mason highlighted the potential for “prompt injection,” in which the AI is given an inappropriate or malicious prompt that causes it to produce content that goes against the intended purpose. This is a key concern for organizations that are looking to leverage the technology, and it is important to have safeguards in place to prevent this from happening.

One of the key challenges with using generative AI is the potential for “hallucinations” – the term used to describe the phenomenon where the AI produces false or fabricated information. This can be a significant issue, particularly in regulated industries like financial services, where it is essential to maintain a high level of trust and accuracy.

The panelists advised organisations on the need to be aware of this risk and take steps to mitigate it, such as training staff on how the technology works and how to identify and address hallucinations. They added that organisations can use tone and structure guidelines to help ensure that the content produced by the AI is consistent with their brand and policies.