EBAday 2024: AI’s Transformative Role in Banking and Payments
I've just returned from EBAday, the annual summit for leading payments and transaction banking executives, held in beautiful Lisbon, Portugal, this year. The Euro Banking Association organizes this annual conference. The organizers mentioned that the event had its highest attendance ever, with 1,300 delegates.
I had the pleasure of moderating the “AI in Banking and Payments” session on the second day. I was joined by Sara Amara, head of currency and clearing Europe at HSBC; Gurumurthy Palani, head of transaction banking at Gulf International Bank; Christian Sarafidis, chief executive of EMEA financial services at Microsoft; and Andy Schmidt, vice president of global industry lead – banking at CGI. EBAday's event partner, Finextra, published an excellent summary of the session, and I wanted to add some observations from my preparatory conversations with the panelists.
The session focused on four central questions, with additional questions from the audience:
- Do you see productivity improvements from traditional AI in your daily banking and payments operations?
- What GenAI use cases do you foresee adding the most significant value? In what areas?
- We hear about new GenAI proofs of concept almost daily. What will it take to scale GenAI across the enterprise, creating new revenue streams and enhancing customer experiences?
- What are the critical risk and governance considerations banks should address when implementing Gen AI?
Productivity Benefits from AI:
Sara Amara kicked off the discussion by saying: “Zero friction is what we aim to achieve for our clients, and the progress that we’ve managed to deliver is what we call predictive and personalized customer service.” HSBC has seen huge success with AI, particularly FX Prompt, a cross-border payment service that informs customers about FX options across currency corridors. Sara believes that the industry needs technology to achieve gains bolstered by the combination of data and AI.
Gurumurthy Palani added a similar sentiment, saying that the sweet spot is a combination of AI, automation, and checking processes. GIB is using AI for trade document checking, fraud prevention, and chatbot deployment. The AI extracts and classifies data, helping to augment the bank’s operations team.
Andy Schmidt highlighted the capabilities of AI in synthesizing data more quickly -- customer service, personalization -- that’s where AI can shine in banking.”
GenAI Use Cases
At the beginning of the session, we launched a poll question -- Over the next year, what will be the most significant use cases for GenAI?
- Uncovering client insights for front-office staff
- Improving customer service (personalized, efficient)
- Streamlining operations (data entry, compliance checks)
- Enhancing risk assessments and fraud management
- Accelerating code generation and testing
The poll results were quite diverse, showcasing the broad potential that GenAI offers financial institutions. 31% responded that streamlining operations would be the primary use case, 29% voted for improving customer service, and 25% voted for enhanced risk assessment and fraud management.
The panelists spoke about how they are working with GenAI. They are looking at payment instructions, onboarding, automating documentation, and improving customer engagement. Many banks are starting at the customer engagement layer, augmenting contact centers and fraud management.
Scaling GenAI
When asked about scaling GenAI, the panel emphasized the need to establish clear goals, alignment with business strategy, leadership commitment, and governance framework. It also takes a scalable cloud infrastructure, with many banks struggling with the data analytics and technology infrastructure maturity required to harness GenAI at an enterprise level. Representing Microsoft, Christian Sarafidis spoke about the speed with which AI is being adopted and developed across the industry. “Chat GPT reached 100 million users in two months, that has never happened before.” He emphasized that another key to scaling adoption is the model itself. Considering 92% of AI use cases are less than a year old, he stressed that at this point of technology adoption, buying out of the box is encouraged rather than building your own solution in-house.
Evaluating the Risks
Lastly, the panel discussed risk and governance considerations. They range from risk of failure – “if you don’t know where you’re going, you’ll never get there”, to varying regulations across global jurisdictions. With the EU AI Act on the horizon, banks must classify the risks associated with AI. For example, credit risk rating, instant payment validation, and verification of payee could all become regulated processes. Technology providers in the EU also must comply with DORA, a set of regulations that require financial services firms doing business in the EU and the businesses that supply them with technology services to improve their digital operational resilience. The DORA Act effectively makes firms like Microsoft regulated entities.
The panelists also discussed how maintaining resilience and protecting data are crucial to maintaining the health of the industry overall. This includes ethical considerations, continuous monitoring and feedback, and maintaining caution as a bank grows its GenAI use case library.