The future of artificial intelligence in the financial sector hinges on a hybrid approach that merges the generative capabilities of large models with the precision of traditional AI, according to Ming Jin, International CTO of OneConnect Financial Technology. Speaking at the Fortune Brainstorm AI conference in Singapore, Jin highlighted the necessity of this strategy for industries like finance, which manage sensitive data and demand high reliability.
Jin noted that the banking sector often grapples with outdated systems, and AI, particularly large language models, offers innovative solutions for modernisation. “By using AI large models to analyse programme code and extract functional specifications, we can quickly understand and reconstruct the underlying architecture,” he stated. This method has proven effective for financial institutions tackling system modernisation challenges.
OneConnect’s parent company, Ping An Group, exemplifies AI’s practical application, having integrated AI across sales, service, operations, and management. In 2024, Ping An’s AI call centre managed 184 billion customer interactions, handling 80% of total enquiries. Additionally, AI-driven anti-fraud technologies have significantly enhanced credit and fraud prevention.
Jin emphasised three critical elements for successful AI implementation: robust data, computing power, and algorithms; professional talent with cross-disciplinary expertise; and a diverse portfolio of business scenarios. He pointed out China’s advantages in these areas, citing Ping An’s extensive resources, including over 3,000 scientists and 21,000 technology developers.
Looking forward, OneConnect aims to advance its AI capabilities, focusing on integrating large models into business processes whilst maintaining human oversight. This approach ensures system flexibility, transparency, and reliability, contributing to the digital transformation of global financial institutions.
“`
