Source: 21st Century Business Herald

At the World Artificial Intelligence Conference 2021, the South China Morning Post reporter learned that financial institutions, represented by banks, are accelerating their digital transformation and speeding up the pilot of integrated application of financial data.

Since the 14th Five-Year Plan proposed to build a new advantage of digital economy, digital economy has become an important driver of economic growth, and digital transformation has become a topic of concern for all industries, especially for the financial industry with strong demand for data.

At the 2021 World Artificial Intelligence Conference, South Finance All Media reporter learned that financial institutions represented by banks are continuously accelerating digital transformation and speeding up the pilot of integrated application of financial data.

At a time when data security is a concern, privacy computing has become the "Netflix" technology of this year's World Artificial Intelligence Conference, which can solve the problem of collaborative computing between untrusting participants while ensuring data security.

This reporter learned from a number of banks and technology service providers that many domestic banks have carried out privacy computing-related research and opened pilot applications.

Finance is a natural scenario for the application of privacy computing technology

Privacy computing is the technology and system of joint computing by multiple participants, and the task of computing and analyzing data is completed by collaborating on joint machine learning and joint analysis of their data under the premise that the participants do not disclose their respective data.

"What is shared among the privacy computing participants is the algorithm, not the data itself, which fundamentally avoids data leakage and technically realizes data security and privacy protection." Yao Ming, founder and chairman of Insight Technology, a privacy computing technology service provider, told reporters in an exclusive interview.

From the technical point of view, contemporary privacy computing includes trusted execution environment, secure multi-party computing, homomorphic encryption, zero-knowledge proof, differential privacy and federal learning, and other system security technologies and cryptography technologies to ensure the security and privacy of the original data while achieving the calculation and analysis of the data.

Yao points out that government and finance are natural application scenarios for privacy computing and are the most widely used areas. For the financial sector, data is an important production material, and the financial sector as a whole has the most complete digital infrastructure and is the most advanced in terms of data diversity, available data volume, and data value development.

However, for the financial industry, how to ensure the compliance and security of data circulation is the most prominent difficulty.

Wu Jincheng, director of Shanghai Economic and Information Commission, said that in the process of accelerating the evolution of artificial intelligence technology, AI technology has brought data privacy, algorithm discrimination, security risks and other issues that have become common concerns around the world, the need to improve the security features of technology and data protection capabilities.

"In previous years, big data risk control has promoted anti-money laundering and anti-fraud in financial institutions, and has helped in both credit collection and marketing of financial institutions." Yao believes that "privacy computing on this basis, making data available invisible, technically reconstructs the data collaboration, data sharing and data circulation of financial institutions."

Financial data comprehensive application pilot launched, many banks layout privacy computing

In May this year, the central bank launched the pilot of comprehensive application of financial data, involving commercial banks, clearing institutions, non-bank payments in Beijing, Jiangsu, Zhejiang, Shandong, Henan, Hubei, Hunan, Guangdong, Guangxi, Chongqing, Sichuan, Guizhou, Gansu, Xinjiang and other places.

In the current financial data circulation and sharing, the South China Morning Post reporter learned that the Bank of Communications, Pudong Development Bank, China Merchants Bank, Weizhong Bank and other banks have launched privacy computing technology research, pilot applications to promote the construction of digital transformation of banks.

During the "two sessions" this year, Cui Yu, a deputy to the National People's Congress and a counsellor of the Central Bank, suggested promoting the safe use of financial data and resource sharing. He suggested that the establishment of improved industry data collaboration governance norms, play the role of financial infrastructure, and explore financial data sharing mechanisms. While accelerating the research on data property rights legislation study, it suggested further strengthening the interconnection of financial infrastructure.

According to Lu Wenjie, senior researcher of Bank of Communications Golden Research Institute, Bank of Communications has been focusing on privacy computing-related technologies since early 2020, and has cooperated with China Mobile, China Telecom, China UnionPay, Hainan Big Data Center, Shenzhen Municipal Government Services Data Management Bureau and other units to carry out pilot applications of privacy computing projects, involving various business scenarios, and has landed a small and medium-sized micro and medium-sized enterprise financing service project based on multi-party secure computing. The pilot application of privacy computing project has been carried out, involving various business scenarios.

According to the reporter's understanding, the current use of privacy computing among banks is mainly in the field of risk control and marketing, but in the future, privacy computing will also play an important role in data sharing among banks and internal organizational synergy of banks, opening up "data silos", realizing cross-border circulation, and becoming an important means of risk prevention for banks.

The banking industry still has the problem of "not knowing how to use" data

Although banks have opened a comprehensive application of financial data pilot, but the data "will not use" problem is still prominent.

Professor Huang Lihua, Secretary of the Party Committee of the School of Management of Fudan University, pointed out that in the practice of public data enabling inclusive finance, resources such as government data and industry data are provided to data service providers through data circulation platforms, and flow to financial institutions after data product processing, and financial institutions enhance their financial services after using data services, so as to export to small and micro enterprises and realize the closed loop of financial data circulation.

However, Professor Huang Lihua found during his visit that some financial institutions currently do not understand the meaning of data after obtaining public data and industry data, nor do they have the ability of data processing and modeling, which will lead to the waste of data resources.

"The efficiency of using data varies widely among the 30-odd enterprises and banks we visited." Huang Lihua mentioned that "some banks use it very well and some banks do not use it well, depending on the ability of the financial institutions as the demand side."

Huang Lihua called for the development and use of financial data, there is still a need to establish a regulatory and evaluation system to achieve a closed loop of financial data value circulation.

South Finance full media reporter learned from a number of industry sources, Shanghai is currently planning to establish the Shanghai Data Exchange in Pudong, or will assume part of the regulatory functions to guide the safe and standardized use of data.