What Risk Officers Need to Know about How Generative AI is Transforming Financial Services

Generative AI is transforming financial services, and Olivier Bessi is considering the opportunities and the risks that this evolution entails

We are witnessing a remarkable transformation in technology and innovation with the advent of generative artificial intelligence (AI), mobile banking and digital wallets. Indeed, the rise of AI has been one of the most significant game-changers in recent years. According to Jones Lang LaSalle’s 2023 Banking and Finance Outlook report, banks globally plan to spend an extra $31bn on AI embedded into existing systems by 2025.

Predictive AI in financial services has been around for some time, enabling financial organisations to streamline operations, improve efficiency and lower operational costs. Generative AI, however, is taking this one step further and is likely to have a transformative effect.

“AI programmes make up facts or sources, leading to misinformation”

Beyond automation, we are seeing exciting new AI opportunities, unleashing previously untapped potential in businesses. Generative AI-driven customer service solutions will be able to adjust to different levels of financial and digital literacy, meet different needs and automate repetitive or routine workloads, allowing humans to focus on specialist tasks such as debt recovery and handling insurance claims.

Managing and resolving issues quicker and more efficiently will help build customer trust and ultimately lead to frictionless service.

Generative AI can also work hand in hand with open banking, allowing lenders to make more accurate decisions based on richer financial data tailored to a particular customer’s risk profile. Recent developments are also democratising AI-enabled trading, opening up a new style of wealth management. At the same time, AI can improve fraud detection methods by enriching transaction data and monitoring new areas of customer activity. Equally, generative AI could enhance customer verification by creating unique security prompts for customers to use while logging in.

It would be remiss to look only at the opportunities that generative AI will bring to financial services while ignoring the risks. So here’s a note of caution; generative AI has its limitations and several risk factors that need to be addressed.

1. Despite recent headlines suggesting that artificial intelligence will replace human skills, AI is unlikely to replace financial services advisors and specialists because it cannot capture the nuances of financial services and market conditions. Trust is crucial within this environment, and while faith can be built through a combination of generative AI and human interaction, it can also be undone by system failures or inaccuracies.
2. Training data sets can create performance issues through biased results. Ongoing supervision, guidance and feedback are essential to enhance employees’ performance and gain customer trust. Generative AI models are also prone to what is known as hallucinations. This is where AI programmes make up facts or sources, leading to misinformation, particularly concerning in an environment where trust is so important. Financial services organisations can mitigate the risks of this by limiting how ‘creative’ the AI model is to ensure that what is received is accurate information.
3. Financial services is a heavily regulated industry in which compliance with strict regulations is critical for companies to maintain this customer trust. This is especially true of cybersecurity, using personal and highly sensitive customer information. Financial organisations have always had to tread carefully between innovation and industry regulation. With the emergence of generative AI, this balancing act is even more imperative. Businesses must fully understand the fundamentals of data security and customer privacy to use the technology safely and confidently without risking their compliance posture.

AI, and particularly generative AI, has the potential to dramatically transform how financial institutions and customers access and interact with financial services. But right now, practice and experimentation within a risk-free environment is essential to truly understanding the benefits and limitations of generative AI in financial services.