Tomorrow’s bank is digital. It’s why we value innovative and creative thinkers in technology who can help us deliver the next generation of banking services.

HSBC's talk and panel will cover technical approaches to develop AI applications in a regulated industry and our views on what the future holds for AI. Data scientists and AI leadership will present, discuss, and take questions from the audience in a panel format.

HSBC has previously co-authored papers on financial document processing to enable language model applications in banks leveraging RAG based methods. This talk will update on the latest developments in testing large language models on document-focused tasks within the bank, aiming to enable chatbot and AI assistant applications. HSBC Office of Applied Artificial Intelligence colleagues will share how we use Generative AI in one of the world’s biggest banks.

Responsible AI is at the core of our ambitions to Digitise at Scale. We will also cover the latest from our research on the human impact of AI.

Key highlights of our approach: •Specificity – LLMs are widely considered a general-purpose technology, however, in a regulated industry, very clear boundaries need to be drawn on what an AI tool can be called upon to do. •Integration – HSBC is a large financial institution operating in 62 markets. Business processes involve many different systems and many can benefit from GenAI, However, regional regulations essentially mandate a one-by-one approach. Any technology being considered needs to be sufficiently available and flexible to cater to these requirements. •Monitoring and Transparency – Any use of AI needs to be constantly monitored and assessed. This of course includes technical metrics of accuracy but, crucially, we need the tool to be transparent in its output. As far as possible, we need to be able to trace how the tool arrived at its output, and further judge if the human actions which resulted were justified and appropriate.