Generative artificial intelligence (gen AI), known for its content creating capabilities, is based on strong learning and generative models, and has emerged over the past year as a key technological advancement which could impact almost all industries.
Among banks, gen AI holds the potential of creating between $200-340 billion in value, according to a prediction from consulting firm McKinsey. Corporate banking, retail banking and software engineering are among the top business sectors that are predicted to benefit.
While AI technologies have offered promising returns, Violet Chung, senior partner at McKinsey, suggested that banks and other financial institutions should be clear about the cost efficiency and productivity gains brought by large scale rollouts.
“There should be top-down considerations as in where gen AI could provide the most benefit,” she told FinanceAsia. “Decision makers should also be informed of possible risks around it.”
Drilling down to specific use cases, Chung cited a report, titled “Capturing the full value of generative AI in banking”, saying that services including coding and sales are seeing the most immediate levers from gen AI, which helps automate processes and providing solutions to more predictive scenarios.
Customer engagement is a one of the creative use cases of gen AI in banking, an example of which is a loan drawdown process made more streamlined and less onerous, the report noted. Technologies such as OpenAI’s large language model GPT-4, which can access images as well as text, is another key application and can potentially help bankers conduct more efficient research as the AI accesses a huge internal knowledge base, and can respond to a wide range of requests quickly. However, while available in most countries, it is currently not available in China or Hong Kong.
To incorporate gen AI in institutional-level banking practices is no easy task, she added.
Infrastructure is needed to accommodate the complicated data gathering and processing capabilities required, while legal and compliance risks should be taken into account. There are the risks of biased or false generation results, intellectual property violations, and potential issues around data security and privacy.
Kanv Pandit, group managing director, Asia Pacific banking services, at fintech service provider FIS, agreed that clearer “guardrails” are needed when gen AI is being incorporated into the financial services sector.
“Content generated needs to be kept from privacy breach, copyright and intellectual violation. It also needs to be non-biased and non-offensive,” he said. Regulators play key roles in providing references, while industry players should be cautious of possible risks at this stage, he said.
The interest in gen AI in 2024 would continue to be strong, he predicted.
“The theme is still leveraging this innovation with clear benefit statements. But institutions need to deploy it in responsible ways, making sure that risks are known and can be mitigated.”
Talent needs
On top of these challenges, a lack of technology talent is a more immediate challenge, said Chung.
Banks need a great number of tech talents, including engineers dedicated to data sourcing, analysis construction, and large language models, among many other specific technologies. A legal and compliance team is also crucial to navigate regulation and risks around AI.
Banks in Asia are slightly better positioned amid the global talent race with larger talent pools available in markets including India and China, she added. Another human-driven advantage of Asian banks lies in consumers, who tend to be more friendly and acceptive towards digital trends.
The demands from a young and tech-savvy population in the region will naturally motivate financial institutions to continue innovation.
Chung added: “When technology innovations are recognised by consumers and the capital market at large, management boards will be much more confident in doubling down their tech efforts.”
¬ Haymarket Media Limited. All rights reserved.
