资源描述
Transforming Paradigms A Global AI in Financial Services Survey With the support of January 2020Transforming Paradigms A Global AI in Financial Services SurveyTable of Contents Forewords . 6 Research Team . 9 Executive Summary .11 Chapter 1: Introduction .16 1.1 A Brief Juxtaposition of AI and Machine Learning .16 1.2 Literature Review .17 1.2. Survey Methodology and Sample Statistics .21 Chapter 2: The Adoption of AI in Financial Services .26 2.1 State and Development of AI Adoption . 26 2.2 Specific Application Areas of AI . . 30 2.3 AI-Empowered Product- and Process Innovation Approaches . . . 33 2.4 Investment in AI . 35 Chapter 3: The Business Impact of AI . 40 3.1. The Future Business Relevance of AI . 40 3.2 How AI Affects Existing Business Attributes . 43 3.3 Propelling Novel Business Value Through AI-Enabled B2B Offerings . 44 Chapter 4: Hurdles to AI Implementation .50 4.1 Overall Implementation Hurdles . 50 4.2 Hurdles for AI Leaders and Laggards .52 4.3 Hurdles Across Financial Services Sectors . 53 4.4 Management Teams Understanding of AI . 54 Chapter 5: Market-Wide Implications of AI Implementation .58 5.1 The Impact on Jobs . . 58 5.2 The Potential for Competitive Disruption .59 5.3 The Impact of Big Tech . 60 Chapter 6: AI as a Risk Driver in Financial Services .64 6.1 The Risk Landscape in an AI-Enabled Industry . . 64 6.2 Reconciling the Market- and Firm-Level Risk Outlook . 66 6.3 Mapping AI-Related Risks by Sector and Jurisdiction . 68 6.4 Risk Mitigation and the Role of AI . 70Chapter 7: Regulation of AI in Financial Services . 76 7.1 AI A Nascent, Global Regulatory Agenda .76 7.2 Beyond the Regulatory Burden . 77 7.3 Supportive Regulation as Comparative Advantage or Under-Regulation as Unfair Advantage? . . 80 7.4 Are Regulations Enabling or Impeding AI Adoption? .81 7.5 Relationship with Law Enforcement . 83 Chapter 8: The Use of Data for AI in Financial Services .88 8.1. The Importance of Data . 88 8.2. Data Sources . 88 8.3. Usage of Customer Data . 89 8.4. Usage of Alternative Data . 90 Chapter 9: Deep Dive AI-Enabled Credit Analytics .94 9.1. Expected Benefits of AI-Enabled Credit Analytics . 94 9.2. Will the Usage of AI in Credit Analytics Exacerbate Bias? . 95 Chapter 10: Deep Dive Investment Management . 100 10.1 Using AI in the Investment Process . 100 10.2 Future Outlook .102 Chapter 11: The State of AI-Enabling Technology .106 11.1 Autonomous AI the Future of Financial Services? . . . 106 11.2 Implementation of Underlying Machine Learning Paradigms . . . 108 11.3 The Use of Computational Resources .110 Chapter 12: Learnings and Outlook . 114 12.1 Generalising Findings Across the Financial Services Industry .114 12.2 Developing AI Capabilities a Must for Financial Service Providers? 11 5 12.3 The Future of AI-Enabling Technology .115 12.4 Future Power Dynamics in Financial Services .116 References .120Forewords 6 Forewords In 1950, five years before the term Artificial Intelligence (AI) was coined by John McCarthy, Alan Turing already posed the question “Can machines think?” and devised the Turing Test. 70 years on, the worlds computational capability has grown by leaps and bounds, and so has the application of AI across a wide array of industries, including Financial Services. However, beyond the news headlines and opinion pieces, there is still very limited empirical evidence available on the current state of AI adoption in finance and its implications. This global survey, jointly conducted by the Cambridge Centre for Alternative Finance (CCAF), at the University of Cambridge Judge Business School and the World Economic Forum, is aimed at going beyond the hype and hyperbole, to provide some empirical data and shed light on the evolving landscape of AI-enabled Financial Services. Based on a survey sample of 151 firms which included both FinTechs and Incumbents, this study was able to depict a global Financial Services sector that is undergoing profound digital transformation underpinned by the advancement in AI. The research findings point to increasing adoption of AI in finance, as firms are leveraging AI to revamp existing offerings and create new products and services. AI is helping firms transform practices, processes, infrastructure and underlying business models, for example selling AI as a service. This research unveils how Financial Services firms are facing hurdles to AI implementation, including access to data, access to talent, and regulatory uncertainties. This study also examined potential and realised risks with growing adoption of AI in finance, the impact on workforces in both the short and long term across industry verticals, and strategic learnings from the current frontrunners of AI implementation. Nevertheless, it is evident that more research needs to be done in order to better understand the opportunities and challenges brought about by the eventual mass adoption of AI in Financial Services. For instance, how can finance firms open up the black box of AI and facilitate more explainable and transparent applications? As AI is becoming increasingly autonomous, what will the roles of humans be and how would an effective human-in-the-loop AI system manifest itself? What are some socio- economic repercussions and ethical implications of AI-induced biases and risks? How can regulators and policymakers harness technology solutions to effectively regulate and supervise AI in finance? This report, therefore, marks just the beginning of a long journey for us to collectively comprehend the potential, possibilities, and boundaries of AI in finance. We are profoundly grateful to EY and Invesco for enabling us to produce this empirical study and for their helpful feedback during the research process. We are also very thankful to the financial service providers who took part in our global survey. Finally, we would like to thank the interdisciplinary CCAF-Forum research team led by Lukas Ryll, which over the last many months worked tirelessly and collaboratively to create this study. Bryan Zhang Executive Director Cambridge Centre for Alternative Finance Matthew Blake Head of Financial driving different business models; underpinning new products and services; and playing a strategic role in digital transformation. The findings also reveal how financial service providers across the globe are meeting the challenges of AI adoption with its emerging risks and regulatory implications, as well as the impact of AI on the competitive landscape and employment levels. The overarching findings of the study suggest that AI is expected to transform a number of different paradigms within the Financial Services industry. These anticipated changes include how data is utilised to generate more actionable insights; business model innovation (e.g., selling AI as a service); changes to the competitive environment with the entrance of Big Tech and consolidation; various impacts on jobs and regulation; impacts on risks and biases; and the further development and adoption of game-changing technologies. The pace of AI application in Financial Services is clearly accelerating as companies begin to leverage AI to increase profitability and achieve scale. This has complicated and multifaceted implications and repercussions. The key findings of this empirical study are as follows: AI is expected to turn into an essential business driver across the Financial Services industry in the short run, with 77% of all respondents anticipating AI to possess high or very high overall importance to their businesses within two years. While AI is currently perceived to have reached a higher strategic relevance to FinTechs, Incumbents are aspiring to catch up within two years. The rising importance of AI is accompanied by the increasingly broad adoption of AI across key business functions. Approximately 64% of surveyed respondents anticipate employing AI in all of the following categories generating new revenue potential through new products and processes, process automation, risk management, customer service and client acquisition within the next two years. Only 16% of respondents currently employ AI in all of these areas. Risk management is the usage domain with the highest current AI implementation rates (56%), followed by the generation of new revenue potential through new AI-enabled products and processes, adopted by 52%. However, firms expect the latter to become the most important usage area within two yearsExecutive Summary 12 AI is expected to become a key lever of success for specific Financial Services sectors. For example, it is expected to turn into a major driver of investment returns for asset managers. Lenders widely expect to profit from leveraging AI in AI-enabled credit analytics, while payment providers anticipate expanding their AI usage profile towards harnessing AI for customer service and risk management. With the race to AI leadership, the technological gap between high and low spenders is widening as high spenders plan to further increase their R however, this data may exhibit shortcomings in quality due to heterogeneous origins. Conversely, payment providers do not seem widely impeded in their AI implementation. This may be because prevalent hurdles, especially data-related ones, may be less relevant to Investment Management Market Infrastructure and Professional Services Payments Deposits and Lending 13% 45% 30% 24% 53% 31% 6% 25% 22% 33% 6% 25% 15% 20% 6% 9% 9% 35% 52% 30% 55% 31% 19% 11% 28% 46% 17% 34% 12% 11% Market uncertainty Cost of hardware/ software Systematic bias in data T echnological maturity Access to data Quality of data Trust and user adoption of AI Access to talentChapter 4: Hurdles to AI Implementation 54 payment providers usage profiles which are primarily geared towards harnessing AI in automation, as opposed to creating new value propositions (as shown in Chapters 2 and 3). For investment managers, access to data represents the largest hurdle, with 52% stating it to be a significant obstacle to AI implementation. This may be attributable to the fact that their most frequently used AI applications are remarkably data-centric, especially AI-enabled data analytics, and using new or alternative forms of data. Trust and user adoption are also shown to be a higher hindrance to investment managers compared to other financial service firms, potentially as investment managers clients may be especially sensitive to issues surrounding algorithmic explainability. Companies active in Deposits and Lending are shown to be similarly impeded by issues revolving around data. They are also more hindered by technological maturity than other sectors, which 25% deem an obstacle. 4.4 Management Teams Understanding of AI In addition to the questions discussed in the previous subsections, the survey also included a free text option at the end. There, respondents could share give their opinion on AI-related aspects which they felt their senior management needed to understand better given their organisations future AI ambitions. The subject voiced most often especially by banks proved to be the prevailing uncertainty around the value proposition of AI. Respondents commented on the importance of identifying AI-driven business cases with attractive Return on Investment (ROI), as well as communicating the potential of AI and enabling factors to senior management. “The impact/value proposition of AI is underestimated. Funding of AI initiatives is too low to be able to prove the value of AI to the business (.).” Senior executive at a multinational investment- and retail bank This snapshot reveals the prevalent uncertainty around AI, especially in incumbent firms. This uncertainty could stem from convoluted corporate structures which inhibit the dissemination of information, meaning Incumbents must establish leaner communication channels with key technology decision-makers, as well as potentially creating new roles geared towards technology for higher executive levels. Respondents also frequently noted the lack of space and resources for AI experimentation. Several participants stated they believed that their company should allow the use of open- source software, offer a sound methodology for developing and testing AI-enabled solutions, and build platforms for model construction and implementation. These concerns reinforce the abovementioned need for technology-oriented roles in senior management, as well as pointing towards the importance of AI sandboxes. As these points were exclusively remarked by Incumbents, they might also provide a clear rationale for creating spin-off entities to establishing less hierarchical and more agile environments, which are more conducive to AI development and testing.Transforming Paradigms A Global AI in Financial Services Survey 5556Transforming Paradigms A Global AI in Financial Services Survey 57 5. Market-Wide Implications of AI Implementation In summary, it is clear that the adoption of AI will bring with it some profound changes to Financial Services. Whilst the technology may drive more job growth in FinTechs, this will be dwarfed by the reduction of jobs in operations and other areas of Financial Services, with an overall 9% anticipated 10-year job reduction in Incumbents, but over 20% in some industry segments. Whilst AI facilitates new and innovative propositions, especially as a core of many FinTechs propositions, the impact on the overall competitive landscape is not expected to be very significant. However, the way that AI technology could be deployed by Big Tech firms, who are in many ways a leading source of AI innovation, is causing great concern amongst Incumbents. Concerns are particularly pronounced in China and the UK while being less prevalent in the US. Key FindingsChapter 5: Market-Wide Implications of AI Implementation 58 Chapter 5: Market-Wide Implications of AI Implementation 5.1 The Impact on Jobs The impact of AI on employment has been much heralded across all industries. One study estimates that over 25% of jobs are at risk due to automation and AI by the end of the 2020s, tailing off thereafter (Hawksworth and Berriman, 2018). The Financial Services sector is expected to be one of those most impacted in the near future. The employment impact of automation and AI on Financial Services is expected to be the greatest of all industries into the late 2020s, with only the transport industry experiencing greater impact in the long term (Hawksworth and Berriman, 2018). The World Economic Forum has estimated that by 2027, 23% of the jobs in Chinas financial sector will either be removed by AI or will be transformed into new positions. The Forum asserts that the remaining 77% of jobs will not be replaced, but the efficiency of these positions will increase, with about 2.3 million people being affected by the impact of AI, that is 23% of the total workforce in the financial sector (He and Guo, 2018). Given the large numbers of people employed within
展开阅读全文