The global hype cycle of AI, driven in large part by ChatGPT, is dying down and real-world artificial intelligence (AI) adoption and application are taking hold. Early adopters are reaping rewards, and AI leaders are driving significant change in their business models.
Banking as a sector was quick to grasp AI's potential, but cloudy and uneven global regulation and examples of initial AI implementation failures have made it harder for some banks to deliver tangible results.
Our recent study with IDC on the use of data and AI in the Asia-Pacific region presents an opportunity to take stock of what’s happening regionally and gain a clearer picture of both achievements and challenges.
AI in APAC banking state of play
As banks work to lower their operational costs, they’ve placed big bets on AI to support automation and deliver efficiency across the front, middle, and back offices. According to the latest SAS and IDC study, Data and AI Pulse: Asia Pacific, almost all banking organizations are highly optimistic about delivering sustainable returns on their AI investments. Some 47% of banks surveyed said they expect returns of up to two times their investment. Another 46% say they anticipate more than three times return.
Are these expectations realistic considering where banks are right now with AI? While use cases and opportunities abound, there are also differing levels of AI maturity, potential challenges with implementation, and significant risks should implementation go awry.
In 2023 and 2024, banks hustled to adopt AI, but their approaches were often scattered. Many banks chaotically adopted untested proof of concepts leveraging disparate technologies and focused heavily on solving productivity challenges. Moving into 2025, we expect banks to pivot toward a more strategic, practical and scalable implementation of AI as they mature in their approach.
Achieving real benefits
Beyond automating existing processes to cut operational costs and increase productivity, banks that hope to achieve the impact they expect from their AI investments must move toward strategic adoption that delivers real business value.
AI is data-hungry, and without a good data foundation, any AI initiative will eventually crumble. Pivoting toward appropriate maturation requires banks to focus on improving their data quality and management and implementing effective governance as the foundations of their success with AI. Banks also must work toward a more tailored approach to their AI technology stack, using fine-tuned, customized models paired with enterprise intelligence architecture.
Focused effort and investment in these areas will position banks for more effective selection and implementation of use cases that can deliver measurable impact and drive business transformation. Transformative AI implementations can help banks better detect and prevent fraud to stop threats before they affect the organization or its customers. AI also could help banks improve loan application processing to reduce errors and deliver faster cash in hand. Or AI can help manage credit risk to reduce the chances of default. Most importantly, AI puts the bank’s strategic goals in the driver’s seat and customers at the heart of every decision.
Addressing AI challenges
While challenges abound, respondents to our study highlighted their top three AI implementation challenges as:
- Insufficient data foundation.
- Lack of clear evaluation criteria for AI solutions.
- Lack of specialized skilled personnel.
Nearly a third of the banks we surveyed cited data challenges as a blocker to progress. They mentioned problems with incomplete or inaccurate data, difficulty accessing data sources, and the high cost of data storage and processing.
Further complicating the data challenge, layers of legacy systems, platforms, and architecture keep data siloed and prevent integration. Lack of integration creates challenges further down the line, as banks miss the opportunity to use the full power of their data to drive strategic decisioning. Breaking down business silos to access and integrate data and layer it with enterprise intelligence is key to delivering real business value.
Another 27% of the banks we surveyed noted no clear strategy or criteria for evaluating AI use cases or solutions. This results in lost time and capital as banks chase one use case after another without effectively prioritizing them. Understanding the bank's overarching strategic direction and goals and applying an aligned and supportive AI strategy is imperative to define the criteria for use case selection.
For example, if one of the bank’s strategic priorities is to deepen and extend relationships with existing customers to prevent attrition, the bank might consider aligning AI use cases that:
- Reduce customer friction including loan application processing, credit review and approval, and fraud management and mitigation.
- Transform customer experience across all touchpoints by improving the speed and efficacy of customer service and transforming the digital experience.
- Support hyper-personalization in real-time, including AI for marketing that provides deeper customer insights, journey mapping and orchestration, and targeted 1:1 next-best offer campaigns.
Finally, at least 35% of respondents noted their banks lack personnel with the specialized AI skills needed to truly transform their business. To address that, banks might consider whether part of the solution to this challenge lies within the challenge itself. If the bank began its AI journey with a focus on automation, efficiency gains, and process improvement, it should have realized enough gains to move employees into new AI-specific roles that deliver greater business impact or, at the very least, freed up enough time to invest in the training, development and upskilling of existing employees.
Delivering the AI-fueled future
Within the next few years, we expect banks to drive further and evolve from adopting and implementing successful AI use cases to becoming true AI-fueled businesses. To make this a reality, banks must integrate AI across the business. They’ll have to solve their data and data governance challenges, build a successful AI strategy aligned to business goals, transform business operations, and start embedding AI into core processes that accelerate innovation.
The AI-fueled bank is more agile and able to adapt quickly to shifting markets, changing consumer demands, and growing competition. The AI-fueled bank will have eliminated internal silos and connected and integrated data so it can flow into a unified decision intelligence platform and deliver data-driven insights that support strategy.
The AI-fueled bank will successfully reshape its customer relationships through improved processes and decisions, resulting in less customer friction, a clearer customer journey, and a better experience based on deep insights. All of these actions and the resulting transformation will allow banks to not only compete but also rise above and deliver the bank of the future.
