Speaker 1: Good afternoon. I'm very happy to be here to go through our financial security lab and what kind of achievements we've made. Just as you have watched in the video, there are four key words. First, the underlying infrastructure, why our lab is building a platform, and why we want to carry out research about financial data security. And the second is key technologies. The key points of current available technologies are the subject of our consideration. And third, the application scenarios. How can we make use of technologies and to use them add value to the market. And now we are serving, that is our objective, in order to better serve the members of the alliance and our customers. And fourth, we have a platform to report, which are going to be shared on this occasion. First, the importance of data. Since the fourth plenary session of the 19th CPC Committee, data has been prioritized as an important factor. And so far, as the instructions made by PBOC in protection of data security and the security of data during circulation, we must integrate the data in different areas. And in the financial sector, data such as the applications, such as customer profiling, optimization of operations, among other areas, data is very useful and fintech can be used. And now we have an application detector platform. At the end of 2019, 30% of the risks of data leakage were actually in the financial sector. So, because the financial sector is of great value, data will pose a huge threat. Globally speaking, from GDPR to the drafts of personal financial information protection and data security law, etc., these documents are very important. And according to these measures, the head of a company will be taking major responsibilities for data leakage. So, data security is very important in our work, and the boundaries of responsibilities, as well as the sharing of data across all the systems, have many pain points for us to address. So, in order to address the barrier of data and make use of data is the key objective. Our lab aims to build a platform to address these pain points. So, in such a context, how can we make use of these core technologies to integrate data? First, privacy protection. It is an important path for us to integrate data and integrate fintech into finance with data. So, with this privacy preserving computing technology, we notice that when we are integrating data and calculating data, data can be protected. That is an important note for us to explore the value of data and make use of it. Based on different distribution types, there are two types. First, distributed, and second,

Speaker 2: decentralized. Distributed is based on PE.

Speaker 1: Then, let's look at federated learning. Horizontal federated learning has many features, repetitive sample sizes, and more. Many banks may have similar types. Let's look at vertical federated learning. Data may be different. For example, in the telecommunications industry, they have different positioning. In the transportation area, they have different features and different data, but they have a huge overlapping of these areas. So, we can use vertical federated learning to address this pain point. Security multiparty computing, SMPC, in application, the parties include the data party, computing party, and results party. This was actually proposed by Academician Yao Qi in 1982. This year, there are a series of specifications to guide the application of SMPC in finance. Confidential computing or learning, that is to do trustful calculation in a trusted execution environment. This core technology is not in the hands of our domestic players, so we still have a long way to go in this regard. Differentiated privacy, DP, the most important way of using it is to add noises in results, calculating results. Theoretically, it can prevent any attacks, but the investment and requirements are high. HE, homomorphic encryption, allows computation on these encrypted tasks directly without decryption. In our analysis, we have tested these six technologies through seven perspectives. These encryption technologies are really cross-disciplinary, and there are many great areas. Based on this reality, the lab, guided by the SFIA, will provide a one-stop service for all of you, from building platforms to integration of data and circulation of data across your flow process. We will provide a one-stop service covering evaluation and testing. Based on these researches, let's look at the application scenarios and some of these demand technologies in the finance sector. Through privacy protection technologies, we can see that while applying them in the financial sector, there are many scenarios, including joint crediting, risk grading, ID verification, anti-fraud, etc. These areas are our target applications. Let's look at the blacklisted sharing. Traditionally, in security service, some of the data was not obtained legally. In the future, the integration of data must make sure that the data comes from legitimate ways. The licensed parties, through privacy protection, we can address the usage of data and the issue of invisibility of data. That is our main work. Let's look at anti-fraud. In current existing scenarios, many institutions have a small sample of data, and the modeling is not very good. Also, there are many issues, such as privacy protection. These anti-fraud models among banks are not communicating with each other. By establishing a third-party lab and a shared platform, these parties can communicate with each other and improve their efficiency, so we can also serve the industry.

Speaker 3: Precise marketing. Via integrating data from different parties, we could understand the interests of our users better and draw the user image in a more precise way. For instance, when you are playing TikTok, you may see some ads. In the WeChat moments, you may also see some ads targeted at you as one of the group targets based on your user behaviors. In the future, we may use this targeted advising to bring more traffic for our clients. Given that, let me move on to our recent work. First is the FinTech data platform. The second is our assessment standards. In the future, we will continue to improve the standards based on our own experience and feedback from our members. We will have FIA standards, and we hope we could be a benchmark entity in this area. Also, we have published two reports. CICT is a third-party institute. We share our reports with all parties, and we hope that research institutes interested can come to us and we could jointly carry out research. This is a FinTech security platform. We hope we could support regulators to provide public services and regulatory innovations, and we also hope we could provide services

Speaker 2: for entities to offer FinTech applications. From these negotiations in the early stage to samples alignment to the end stage, we will take security as an extra element and provide technical support in an all-round way. This is the assessment system. We will try to materialize those assessment standards,

Speaker 3: not only in Shanghai but to cities across China, and that is our overarching plan.

Speaker 2: These are the two reports that we have published. Of course, there is still

Speaker 3: room to improve, and I welcome your comments and suggestions. Thank you so much for your listening, and we hope our FinTech lab can have more interactions with all of you, and we believe under the guidance of Shanghai FinTech Industry Alliance, we will continue to support and provide you well-stocked services. Thank you.