AI and financial services: three areas for executives to watch

AI and financial services: three areas for executives to watch

Karen Oakland, Vice President of Financial Services Marketing, Smart Communications, speaks to us about AI, how ChatGPT benefits the industry, as well as automation is heading and some potential risks that may emerge.

Artificial Intelligence (AI) is a hot topic these days and people are increasingly talking about OpenAI, ChatGPT, GPT-4 and large language models (LLMs). AI-powered services have steadily been improving in the background in recent years, but the release and rapid adoption of ChatGPT means that leaders in banks and financial services institutions are looking harder than ever for ways to accelerate its use across the business.

At a high level, most executives are focused on three specific areas: driving revenue and loyalty by improving communications, improving operational efficiencies by eliminating manual work and reducing risk, especially around fraud and security. Let’s take a closer look.

Improving the customer experience

Many bank websites now have AI-powered chatbots that help guide customers to answer questions or to help them make better financial decisions. These solutions are getting even smarter as they can process and use larger data sets, ultimately providing more of a two-way, human-like conversational digital experience available 24/7 from anywhere. 

By connecting them with owned data, like transaction history and demographic information, as well as third-party data like web visits, these chatbots can generate responses with a higher level of personalisation than ever before, providing customers with the type of banking experience they expect on consumer services like Amazon or Netflix. This can ultimately drive ‘stickiness’ with impacts on both loyalty scores and overall wallet share.

For wealth management and other advisory firms, this is particularly powerful. ChatGPT can be used to help advisors with financial data analysis so they can make smarter recommendations to clients, faster. Tools can suggest ideas, build financial models or create pitch decks and author memos. Indeed, Morgan Stanley is currently piloting GPT-4 to help advisors more easily answer questions and perform other tasks. 

Increasing speed and efficiency

Generative AI is having a massive impact on the speed at which content can be created. Contact centre representatives and financial advisors can use AI to jumpstart quicker drafts of correspondence, saving time and improving client engagement. AI-powered chatbots and virtual assistants are capable of handling routine customer inquiries and requests quickly and accurately, freeing up human agents to focus on more complex customer needs. This not only speeds up response times for both humans and AI but also reduces wait times and improves overall customer satisfaction.

Complaints handling is another potential use case in the banking industry. USAA Bank is using a combination of AI-powered natural language processing and manual tagging to manage the company’s response to consumer complaints, according to a presentation at the Consumer Banking Association meeting. The bank is testing the use of Machine Learning and AI-driven text recognition and analysis tools to automatically categorise text and then route complaints accordingly. 

Reducing fraud risks 

AI technology can act as a guide and almost a predictive source to help financial institutions automatically identify fraud or security breaches. AI technologies can analyse transaction data and user behaviour and then identify patterns and anomalies that indicate fraudulent activity. This helps financial services organisations quickly detect and prevent issues, saving both customers and the organisation time and money.

Watching out for pitfalls

It must be noted, however, that utilising AI in financial services also comes with potential pitfalls. Customer communications are highly regulated in financial services, with oversight in the UK from Financial Conduct Authority, among others. Even though AI can help content be produced faster, financial institutions still need to make sure these communications are compliant and auditable. That’s why a structured and secure software solution for this purpose is the best option.

In addition, advisors must be aware that not all AI information is accurate. Regulated industries must proceed with caution due to the risk involved with allowing inaccurate or unaudited content to be created and shared with customers. Inaccurate information, known in the AI space as ‘hallucinations’, can cause a poor customer experience, loss of money and impact a customer’s trust in a brand. If a company provides the wrong content or makes the wrong recommendation, it can ultimately lead to reputational damage and make them vulnerable to lawsuits and regulatory fines.

Moreover, AI systems can also be biased. Machines have trouble balancing diversity and equity factors, so advisors must be aware of the data sets their AI is using to make recommendations and step in if they detect bias. Some argue that the worst AI risks are the ones that cannot be anticipated at this very moment. It might not be long before a financial advisory firm gets into regulatory trouble because its AI-generated communications are biased, inaccurate or possibly recommending actions that result in losses or other unforeseen consequences.

Traditional banking chatbots and static information may soon seem antiquated in comparison. For now, it is best to limit the use of ChatGPT to tasks where the output could be reviewed by a domain subject matter expert. But soon, AI capabilities are likely to improve such that they are capable of these more complex tasks, improving the customer experience and driving even more operational efficiencies.

Click below to share this article