2024重塑生成式 AI 在银行业的角色洞察报告：弥合炒作与现实的差距

Bridging the gap between hype and reality Contents Executive summary GenAI demystified: a sneak peek at its role in bankingRisks in AI adoption: moving with cautionNavigating the regulatory maze: a compliance journey in AI Chapter 1: Balancing hype, risks andregulatory realities A holistic dimensional approach to strategic scaling of GenAIAchieving value from generative AI implementation Chapter 2: From POC to enterprise-wideAI implementation Chapter 3: Adoption framework forscalling GenAI in banking Key considerations for scaling generative AI in the banking industryNTT DATA’s GenAI banking framework 5 Ways financial institutions are already deploying and benefitingfrom AI & ML Chapter 4: Practical use casesbeyond POCs Conclusions Bibliography Executivesummary AI in banking is not new, but it is changing Generative AI is pioneering the futureof banking For a while now, the realm of Artificial Intelligence (AI)hasn’t ceased to surprise, with Generative AI (GenAI)starring the scene since the beginning of 2023. Bankingleaders have hopped onboard, launching numerous pilotprojects throughout the year. Initially, banks have rightlyfocused on enhancing productivity with their GenAI pilots.However, this technology has the potential to significantlytransform job functions and customer interactions,potentially leading to entirely new business models. As banks embark on this journey,adaptability and reinvention areparamount to reap the benefits in theever-evolving terrain of technologicalinnovations and regulatory shifts Gen AI brings new opportunities but also entailsnew risks. Risk management for GenAI is still in theearly stages for financial institutions, melded by theuncertain future of legal obligations, which may varyacross geographies. Nevertheless, sooner rather thanlater, banks will need to activate some no-regret movesto get ahead of impending legal changes. Bankinginstitutions must understand their legal responsibilitiesand accordingly redesign their AI risk, governanceframeworks and controls, resulting in a dual approachrequiring a delicate balance between innovation andrisk management. If handled correctly, it positionsorganizations at a competitive advantage in the rapidlyevolving regulatory landscape. Ultimately, to meetthe expectations of investors, regulators, and otherstakeholders, equals trust. Without losing focus on distinguishing betweenbuzzwords or ‘hype’ and genuine value-adding innovation,the pressing questions for banking institutions now arenot only how and where to use GenAI most effectively,but also how to ensure the applications are fully adoptedand scaled within their organizations. A methodicaland strategic approach is necessary to successfullyintegrate GenAI into banking institutions. This involvestransforming the operating model and customerexperience, fostering an engaged and curious workforce,securing data, and implementing proper guardrails. Thiswhitepaper presents NTT DATA proposed GenAI adoptionframework, designed to empower banking organizationsnot only to keep pace but to yield enterprise-wide AI inthis fast-moving, groundbreaking field. Hype aside, some transformational GenAI innovationsare starting to emerge. Chapter 1:Balancing hype, risks andregulatory realities GenAI demystified: a sneak peek at its rolein banking The banking industry stands to experience one ofthe most significant impacts from generative AI,potentially transforming a substantial percentageof its revenues Banking institutions are increasingly recognizing the vastbenefits of GenAI across the entire value chain. This technologyhas the potential to deliver an additional $200 billion to $340billion annually, equivalent to 9 to 15 percent of operatingprofits, largely from increased productivity.1 The pressing questions for banking institutions have shiftedfrom how and where to use generative AI most effectively, tohow to ensure these applications are fully adopted and scaledwithin their organizations. Rarely has the pace of evolution ofa new technology been so rapid. This has inevitably ignited asense of urgency within organizations, resulting in a frenziedrace to develop this type of AI. Differencing noise from music In implementing GenAI into banking institutions, acomprehensive and orderly approach is paramount,considering the complexity and breadth of tasks involved.Correct implementation is intricately linked to avoiding theexaggeration of its capabilities or falling prey to the ‘hype’phenomenon, which can pose significant risks to businesspractices. To begin, it’s essential to dispel the notion of FOMO (Fear ofMissing Out) surrounding GenAI. While it holds transformativepotential, GenAI is not the silver bullet to every banker’sproblem. Taking a breather from chasing trends allows forgaining perspective and revealing how emerging technologiesconverge and connect over time, ultimately fueling betteractions in the present. It’s essential to pause amidst the hype and thoroughly