With the advantage of data technology, digital finance extends from mastering data of commodity flow, capital flow and information flow to financial core business fields such as payment, financing and investment, and uses digital systems to transform many complex and changeable financial information into measurable numbers and data, and then establishes corresponding digital models with these numbers and data and transforms them into binary codes, which can make financial work more agile, have stronger competitiveness in rapidly changing market business and further strengthen financial services.
Digital Financial Technology Elements Surge in "Five Great Articles"
Digital finance refers to a new generation of financial service mode that combines traditional financial service formats through the Internet and information technology. From the definition, the elements of digital finance have the following characteristics: First, digital finance is a digital process of financial industry with financial technology as the core, and it is a revolutionary innovation in the field of financial technology. Financial technology is based on a series of technological innovations such as big data, cloud computing, artificial intelligence, blockchain, etc. It is widely used in six major financial fields, including payment and settlement, loan financing, wealth management, retail banking, insurance and transaction settlement, and is the mainstream trend of the financial industry in the future. Second, the participants of digital finance are diversified. It includes not only traditional financial institutions, but also emerging market entities such as digital platform enterprises and financial technology companies. Among them, traditional financial institutions are the most active, digital finance has promoted the efficiency of financial services, and qualified financial institutions, especially commercial banks, are accelerating the pace of digital transformation. Third, we attach great importance to the development and application of data elements. Financial institutions mainly rely on digital technology, digital channels and digital infrastructure to realize the supply of financial products and services, and data is the cornerstone of digital finance and the foundation of digital transformation of financial institutions. Therefore, financial institutions attach great importance to the development and application of data elements. The digital transformation of financial institutions has basically realized the data support business, and also introduced social security, provident fund, taxation and other data for customer acquisition, marketing, credit granting and risk control. In the future, financial institutions may need to pay more attention to the empowerment of data, that is, the third release of data value.This is the key to promote digital innovation in financial products, services, business processes and business models.
Let’s take a look at the relationship between digital finance and the "five big articles" of technology and finance, green finance, inclusive finance and pension finance. Technology and finance, Green Finance, inclusive finance and Pension Finance respectively focus on supporting financial services and products in the fields of scientific and technological innovation, green and low carbon, people’s livelihood and the elderly, while digital finance uses digital technology or financial technology to promote the innovative development of financial business. That is to say, technology and finance, green finance, inclusive finance and pension finance can form a brand-new business development model as long as they are blessed by digital technology, that is, they will become digital technology and finance, digital green finance, digital inclusive finance and digital pension finance. These brand-new business development models are all digital financial models supported by digital technology, and create the following three application values: First, capillary value. With the help of Internet of Things, Internet, big data and other digital technologies, we can reach more entities, especially small and micro enterprises, enhance information acquisition, storage and application capabilities, and reduce systemic risks. The second is the long tail value. With the help of digital technologies such as artificial intelligence, the effective information in the data is mined, and the matching resources are adjusted based on the value to meet the diversified financing needs, link needs and wealth needs of the long tail Pratt & Whitney group. The third is accurate value. Comprehensive use of digital technologies such as ABCDEFG (that is, A is artificial intelligence, B is blockchain technology, C is cloud computing technology, D is big data, E is experience improvement, F is sensory interface including face recognition, and G is genetic technology), and an agile service platform is formed through digital transformation, which turns passive into active.Finally, the ultimate satisfaction of the diversified needs of the entity enterprises and the general public will be realized.
The big financial model is building a new ecology of digital finance.
The White Paper on the Ecology of Financial Technology in China (2023) recently released by China ICT Institute points out that the technological changes of large-scale models are continuously evolving and bring new opportunities to the financial industry. The financial industry has a good digital foundation and intensive artificial intelligence application scenarios, which also provides a sufficient environment for the application of large models. The so-called financial big model refers to the machine learning and multi-modal artificial intelligence (AI) model with a large number of parameters and complex structure applied in the financial field. As shown in the figure, in such an ecological framework, whether it is a general big model or a vertically applied big model, or even financial big data, computing power and algorithms, they are all components of financial technology. The so-called "deep roots" can be "strong" and "numerous branches" can be "flourishing". All parts are interdependent, mutually reinforcing and closely coordinated, and they can jointly promote the orderly operation of financial big models and further promote the development of digital finance. In addition, the big financial model is also closely related to financial technology and digital finance. The financial big model is a composite technology product that integrates many financial technologies such as big data, cloud computing, blockchain, metauniverse and artificial intelligence. It can be widely used by financial institutions in intelligent risk control, intelligent investment and research, intelligent customer service and other fields, providing them with intelligent and efficient solutions. Moreover, by using the financial macro model, financial institutions can better understand customer needs, enrich business scenarios, enhance customer experience, predict market trends, optimize product design and business processes, and strengthen security and compliance, thus providing strong support for the digital and intelligent transformation of the financial industry. Specifically,In the business front desk, the big model can be used as intelligent customer service to improve the work efficiency of business personnel in outlets; In the business center, the analytical ability of the big model can be used as investment research and investment to improve the accuracy of credit; In the business background, the big model can improve the risk control ability and reduce the investment risk. The financial macro-model can almost run through the whole process and task of the financial industry, and reshape the production relations of the financial industry through the release of the productivity of each endpoint. These innovations fully meet the requirements of the new ecology of digital finance.
On November 23, 2023, Tencent Cloud, as the leader of the domestic financial cloud market, officially released the big model solution for the financial industry, committed to making every financial institution have its own big model. Commercial banks are also actively deploying AIGC and financial models. According to the mid-year reports of listed banks in 2023, nine banks, namely Industrial and Commercial Bank of China, Agricultural Bank of China, Bank of China, Bank of Communications, China Merchants Bank, China CITIC Bank, Industrial Bank, Huaxia Bank and Zheshang Bank, have clearly stated that they are exploring the application of the big model. Among them, Agricultural Bank of China has taken the lead in launching ChatABC, and it has been iteratively upgraded; ICBC takes the lead in realizing the application of tens of billions of basic models in many scenarios, such as knowledge operation assistants and financial market investment and research assistants. Bank of Communications formulates the construction plan of generative artificial intelligence and sets up a special research team of GPT large model; The Postal Savings Office actively explores the research of cutting-edge technical fields such as digital employees, NLP dialogue robots and pre-training models. Recently, it officially released the i-PSBC digital postal savings and four "I" series of digital finance achievements including mobile banking 9.0; China Merchants Bank has improved its ability to build large models, and focused on exploring its application in the whole process of wealth management, and put into production the FinGPT Creative Center. At the same time, all financial scenarios of commercial banks are exploring to adapt to the big model interface, and their procurement needs range from computing power reserve to model procurement, from cloud computing resources to data governance, covering all kinds of manufacturers in the artificial intelligence industry chain. Within the major insurance groups,Bidding for GPU server procurement and data governance service projects prepared for large models is also in full swing. Looking at the digital finance of the whole financial industry, it has entered the era of financial model.
The focus of China’s digital finance innovation and development in the next stage
How to focus on the big article of digital finance, the key is to find out the pain points or difficult problems encountered in the development of digital finance at present, and then break through and solve them. After in-depth practice and careful investigation, the author believes that the pain points or difficulties encountered in the current development of digital finance mainly include the following aspects: First, data security and privacy protection. Digital financial services are highly dependent on the Internet and digital technology, involving the collection, transmission and storage of privacy and personal data of financial institutions and a large number of users. Once a security incident occurs, it may cause great losses to financial institutions and users. Therefore, how to ensure the security and privacy of these data and prevent data leakage and abuse is an important challenge. The second is the inclusiveness and inclusiveness of financial services. At present, there are still some vulnerable groups who cannot enjoy the dividend of digital financial services. How to further expand the inclusiveness and inclusiveness of financial services and let more people benefit from the development dividend of digital finance is also an urgent problem to be solved. Third, the application of the big financial model is "acclimatized". The application of the big model is not smooth sailing, mainly because the vertical application of the big model by financial institutions can not be seamlessly connected with the general big model of the financial industry. The reason is that the digitalization ability of financial institutions is uneven, and the demand for large models is different. In addition, the financial industry generally has strict requirements on digitalization efficiency, security and risk control. Therefore, the big model of the financial industry faces a series of challenges such as low adaptability, lack of scene traction, and frequent "AI illusion".So that the large model can’t land and dock smoothly. The fourth is the ethical issue of financial technology. Financial technologies such as intelligent risk control, intelligent investment, intelligent customer service (robot customer service) and other financial applications based on artificial intelligence and big data technology have almost become the standard of all financial institutions. However, there are ethical problems in the application of these technologies, especially the ethical problems in the application of artificial intelligence technology in the financial field. For example, intelligent algorithm may cause problems such as algorithm black box, algorithm discrimination, credit supply discrimination, and ethical problems such as inducing excessive borrowing. Improper response may affect financial stability and security.
By fully understanding the challenges and risks faced by the development of digital finance, we can better cope with and solve the problems. First, focus on strengthening data security and privacy protection. With the accelerated development of digital finance, data security and privacy protection have become key topics. Strengthening data security and privacy protection can be strengthened from the following aspects. The first is to strengthen data encryption. For important data, use strong encryption algorithm to encrypt it, so as to ensure that only authorized personnel can read and process the data. At the same time, it is necessary to adopt anti-leakage technology and add multiple security layers in the process of data transmission and storage to avoid data loss. The second is to strengthen the management of data rights. Only authorized personnel can access and process sensitive data, and strict access control mechanism needs to be established. This can be achieved by establishing a data center or cloud service. At the same time of authorization, it is necessary to establish a detailed logging mechanism in order to track everyone’s access to data and operation records. The third is to strengthen the protection of personal privacy information. Personal privacy information should be anonymized or desensitized to prevent personal sensitive information from being illegally obtained and used. The fourth is to strengthen anti-fraud efforts. Adhere to the rule that "what you see in online channels is not necessarily true", and be wary of criminals using AI synthesis technologies such as "onomatopoeia" and "face changing" to easily disguise themselves as others, and accurately implement fraud by quickly screening target groups and customizing fraud scripts. Illegal behavior; When transferring money and remittances, people who call themselves "acquaintances" and "leaders" induce remittances for various reasons through social software and telephone calls.Be sure to verify the identity of the other party. The fifth is to establish a complete security risk management measure. Including the establishment of an anti-fraud alliance with the participation of regulatory authorities, industry associations, financial institutions and technology enterprises; Strengthen the protection of consumers’ rights and interests at the industry level, and jointly establish a customer rights protection center to form an industry risk mitigation and mutual assistance mechanism; Establish a sound data security governance organizational structure and staffing, strengthen system construction, and pay attention to personnel management and process construction; Establish a sound data security risk assessment and monitoring mechanism. At present, the promulgation and implementation of People’s Republic of China (PRC) Data Security Law, People’s Republic of China (PRC) Personal Information Protection Law, and the Central Committee of the Communist Party of China the State Council’s Opinions on Building a Data Infrastructure System to Give Full Play to the Role of Data Elements have provided legal protection for data risk prevention and control in the financial industry.
Second, focus on strengthening the inclusiveness and inclusiveness of digital financial services. Digital finance can create three application values: capillary, long tail and precision, and these three application values are fully reflected in the financial services of inclusive groups. To this end, the State Council’s Implementation Opinions on Promoting High-quality Development in inclusive finance (Guo Fa [2023] No.15) requires strengthening science and technology to empower inclusive finance, supporting financial institutions to deepen the use of Internet, big data, artificial intelligence, blockchain and other scientific and technological means, optimizing inclusive finance’s service model, improving credit approval and risk management models, and improving the availability and quality of financial services such as small and micro enterprises, individual industrial and commercial households and agricultural subjects. Promote the development of Internet insurance norms, enhance the ability of online underwriting claims, and improve the level of insurance services through digital and intelligent management. We will explore the pilot project of blockchain construction in the regional equity market in a steady and orderly manner to improve service efficiency and safety management. It is also necessary to create a healthy digital inclusive finance ecosystem, support financial institutions to rely on digital channels to connect with online scenes, and provide high-quality inclusive finance services in the fields of small and micro enterprises, agriculture, rural areas and people’s livelihood. On the premise of ensuring data security, financial institutions are encouraged to explore information collaboration with small and micro enterprises, core enterprises, logistics and warehousing and other parties in the supply chain to improve the efficiency of supply chain financial services for key groups in inclusive finance. Encourage the organic combination of digital government, smart government and digital inclusive finance, promote the convenience of financial services closely related to daily life, and at the same time ensure the daily cash use of the people.
Thirdly, efforts should be made to strengthen the multi-party cooperation in the application of financial big models. From the application practice of financial big model in 2023, the inherent uncertainty and high training cost of new technology determine that it is difficult to promote its landing application by relying on a financial institution’s own technology and R&D strength. Therefore, the author believes that the joint creation and ecological co-construction of financial institutions and technology companies is undoubtedly the most effective way to accelerate the application of big models in the financial industry. As the frontier field of digital and intelligent application, the financial industry is rich in data resources and has a wide range of practical scenarios, which has the prospect of deep integration with large models. Technology companies also need to improve their large-scale model capabilities and explore commercialization paths through the tempering of business scenarios. The cooperation between the two parties has great prospects. From the real case, Tencent Cloud has cooperated with many financial institutions in the application of financial big models in the past year, and achieved positive results. In terms of products, Tencent Cloud TCE has helped a large number of large financial institutions such as Bank of China, China Construction Bank, People’s Insurance Company of China and China UnionPay to build a secure and compliant full-stack private cloud, which is the largest proprietary cloud solution with the largest number of cases in the industry. The advent of Tencent Cloud’s big model solution for financial industry is expected to promote the big model to take root in China’s financial industry. For the "AI illusion" and the risk of robot customer service, it is also necessary to strengthen cooperation, and mainly adopt artificial agent (AI Agent) to effectively solve it technically. AI Agent is different from traditional artificial intelligence.It is an intelligent entity that can perceive the environment, make decisions and execute actions through independent thinking and calling tools. At present, AI Agent has a unique way to solve the problems of big model "AI illusion" and context capacity limitation, and financial institutions may wish to try it in cooperation with technology companies. Looking forward to the future, AI Agent will make the big model become an agent with independent thinking, decision-making and execution ability with the help of one or more agents, and continue to realize the road to artificial general intelligence (AGI).
Finally, efforts should be made to strengthen the ethical governance and construction of financial technology. The so-called scientific and technological ethics refers to the values and codes of conduct that need to be followed in carrying out scientific research, technological development and other scientific and technological activities. To promote financial science and technology to be good, it is urgent to strengthen the ethical governance of financial science and technology, improve the regulatory rules of financial science and technology ethics, strengthen the role of ethical review of financial science and technology, effectively prevent and control the ethical risks of scientific and technological activities in the financial field, and realize the benign interaction between innovative high-quality development and high-level security of financial science and technology. On October 9, 2022, the People’s Bank of China officially issued and implemented the Guidelines on Science and Technology Ethics in the Financial Field, which is an important milestone in building an ethical governance system for financial science and technology. The guideline provides seven values and codes of conduct that should be followed in carrying out scientific and technological activities in the financial field, namely, innovation, data security, inclusiveness, openness and transparency, fair competition, risk prevention and control, and green and low carbon. On October 8, 2023, the Ministry of Science and Technology, together with ten departments, issued the Measures for Examining Ethics of Science and Technology, which tied the "safety belt" and put the "steering wheel" for the accelerated breakthrough of new technologies such as AI and life sciences in China. With the publication of the above-mentioned policy documents on ethical governance of science and technology, the ethical construction of financial science and technology has been integrated into the risk management and scientific governance system of financial institutions, and has become an important part of it. The ethical education of science and technology, information security and data security risk management and control have been strengthened. In the future, the related requirements and protective measures of ethical governance of financial technology should be further strengthened, the ethical governance system of financial technology should be continuously improved, and the governance measures should be fully implemented.In particular, commercial banks should speed up the establishment of relevant organizations responsible for the review of scientific and technological ethics. At the same time, it is necessary to embed ethical governance into the design and implementation of financial technology products and services, give full consideration to language, culture, gender, age and other factors, and provide "temperature-sensitive" services in an intuitive and concise way, so as to avoid inconvenience and obstacles to the "most disadvantaged" and effectively improve the availability, ease of use and security of financial services. In the cooperation between financial institutions and technology companies, we should adhere to the principle of "finance is the foundation, technology is the instrument", adhere to the positioning of technology as financial empowerment, delimit the cooperation boundary between financial institutions and technology companies, and effectively isolate financial risks from technology risks.
(The author is a professor and researcher at Zhengzhou Training College of the People’s Bank of China)
① People’s Bank of China. Financial Science and Technology Development Plan (2022-2025) [EB/OL]. http://www.pbc.gov.cn/goutongjiaoliu/113456/113469/4438627/index.html, 2022-1-4.
② People’s Bank of China. Guidelines on Science and Technology Ethics in Financial Field (JR/T0258-2022) [EB/OL]. http://www.czifi.org/newsinfo/4509335.html, 2022-10-09.
