Viewpoint: The privacy computing explosion is a challenge and an opportunity

What exactly is “privacy computing”? What role can it play in blockchain and the future digital economy?

Author | Ming Feng

In the current exploration of the frontier of blockchain industry, the figure of “privacy computing” has also started to appear gradually.

At present, both BAT and other large companies, as well as technology start-ups, are entering the privacy computing market. Ant Financial Services, Tencent Cloud, and Baidu have launched their own products, and a number of startups focusing on the productization of privacy computing, such as Huajian Qingjia and Fudian Technology, have also emerged.

What exactly is “privacy computing”? What role can it play in blockchain and the future digital economy? What are the challenges to be solved or to be faced by privacy computing?

Privacy computing is on fire
Privacy is a sensitive topic for individuals or businesses. In recent years, discussions on topics such as “Big Data Killing”, “Feeling like your phone is listening to your consumer preferences”, and “Delivery boys trapped in algorithms” reflect the growing need to protect The demand for data security and personal privacy is growing.

So, what is privacy? How can privacy computing achieve the purpose of privacy protection?

Privacy Computing is a technology and system in which two or more participants jointly compute, and collaborate to perform joint machine learning and joint analysis of their data without revealing their own data.

The participants in privacy computing can be either different departments of the same organization or different organizations.

The main purpose of privacy computing is to make data “available and invisible” in all aspects. Compared with traditional data confidentiality methods, its most innovative highlight is the ability to achieve physical decentralization and logical centralization of data, ensuring data security and privacy while also mining data value and promoting value circulation.

As an interdisciplinary and comprehensive technology, privacy computing involves many related concepts: multi-party secure computing, trusted hardware, federated learning, differential privacy, blockchain, etc.

The current mainstream technology routes in the industry include three categories: Federated Learning (FL), Secure Multiparty Computing (SMPC), and Confidential Computing (CC)/Trusted Execution Environment (TEE).

Privacy computing also made the list of nine key strategic technology trends to dig deeper into in 2021, published by Gartner, the world’s leading technology research and analysis firm.

Blockchain and privacy computing need to be combined
Industry insiders generally agree that blockchain and privacy computing need to be combined and empowered by each other to be effective. Blockchain solves the problem of decentralization and trust, while privacy computing ensures that the privacy of individuals as well as commercial organizations is not leaked on the chain, and the combination of the two is the necessary way to build a decentralized society in the future.

The Tencent Privacy Computing White Paper specifically mentions that although privacy computing achieves privacy protection for input data in the process of multi-party collaborative computing, the original data, computing process and results face verifiability problems.

Blockchain, with its technical features such as shared ledger, smart contract and consensus mechanism, can realize on-chain verification of the original data and on-chain retrieval of key data and links in the calculation process, ensuring the verifiability of the calculation process.

Therefore, applying the credible proof of blockchain technology for computing to privacy computing can enhance the verifiability of the privacy computing process while protecting data privacy.

On the contrary, it is well known in the blockchain industry that the biggest strength of blockchain is the absence of centralized management and ownership. How can users be trusted and involved in managing their own data? Can users be reasonably incentivized?

“A strong data security system needs to be attached to the blockchain to establish a closed loop of security where the entire flow of data is under the protection of the data security system without any security gaps, ensuring that no one outside the user (including the project owner) can peek into the user’s data.” Wang Donglin, international cryptography application scientist and founder of YottaChain distributed storage public chain, said to Chain New.

Wang Donglin stressed that blockchain itself does not guarantee data privacy, blockchain uses cryptographic hash and signature but does not contain encryption, and data privacy is achieved by attaching a specific data security system outside the blockchain. The incentive system, on the other hand, is one of the main values of blockchain, but whether the incentive is reasonable depends on how the project is specifically designed and how it is operated.

According to Wang Donglin, “It can be said that reasonable incentives are feasible, but most of the projects cannot do it.”

So in the blockchain industry, privacy computing is also seen as a sure path to Web 3.0. the key attribute of Web 3.0 is that the ownership of data belongs to the users themselves. Privacy computing is one of the very few technologies that can provide multiparty data federated computing that allows data to be executed and protected within a secure environment.

Landing on the ground in progress
After a technology diffusion and market education phase in 2019 and a large-scale proof-of-concept and pilot deployment phase in 2020, privacy computing will begin to enter the real trial scale application phase in 2021.

In the medical field, for example, patient data is sensitive, and various scientific studies based on medical data usually require a large number of samples, and it is difficult for the data volume of a single data source to meet the massive data demand, while the process of data sharing brings the risk of privacy leakage. The application of privacy computing technology for multi-party collaboration can effectively prevent the leakage of medical data associated with sensitive information and ensure data security.

Currently, several medical institutions have jointly constructed a target detection model through a horizontal federated learning solution without the data of each medical institution out of the domain, resulting in a significant increase in effective training data. It is reported that the performance of a multiparty federated training model is more than 30% higher than that of a model trained by a single medical institution.

In the financial industry, Fengxiao Huang, partner and senior director of Rich Digital Technology, has told the media that privacy computing technology has the potential to become the boundary between private domain traffic platforms, and it has the potential to reconstruct the internal logic and business model of the traffic business.

As a simple example, the number of bank deposit users is huge, but it is difficult to circle the users who are interested in purchasing wealth management by simply relying on the customer data in the bank. The general operation is to procure tags from third-party data companies through APIs or offline libraries, and API query actions can largely pose a threat to bank customer ID privacy. Privacy computing, on the other hand, can technically ensure the privacy data security of such operations.

Through the analysis of the case can be found, the current privacy computing technology is mainly used in finance, Internet, government, communications, medical and other fields, the main application scenarios are precision marketing, financial risk control, medical and health, identity verification and so on.

“Privacy computing has been combined into the field of blockchain technology for about a few years, but in the past, because the application scenarios were not highlighted and the pain points were not clear, it has not been able to find the best value play in the implementation of actual business scenarios.” Yuehua Wang, partner of DDT Innovation, admits that privacy and security are the most basic existence in blockchain protocols, but since the actual landing scenario of distributed data was not highlighted in the past, privacy computing was lacking. “The value of data is unlimited, but the only way its value can be reflected and played is if the privacy and security of the data are guaranteed accordingly.”

A challenge, but also an opportunity
There is no doubt that privacy computing itself has great value and application prospects, but there is still a very long way to go to reach the tens of billions of market scale and break through the bottleneck of commercial development.

On the one hand, the current market awareness and recognition of privacy computing is still insufficient. Because privacy computing technology is complex and often “black box”, most users have difficulty understanding and trusting privacy technology. If the technology is not fully understood, users may have excessive expectations of the effect of the technology application.

Take “blockchain + digital identity” industry application as an example, “the current landing effect is general”, a senior product manager of blockchain industry told “Chain New” that the main development bottleneck lies in: firstly, the blockchain technology and data privacy protection technology have the problem of low performance; secondly, the user has the problem of low performance. The second is the conflict between user privacy protection and the business model of enterprise data realization. Under the premise that secure multi-party computing is very immature, the only way to protect user privacy is to authorize identity verification, which cannot authorize identity information, which may lead to no incentive for enterprises to participate; the third is that DID and private key are difficult to remember and need to be saved by users themselves, which has a high threshold for users.

On the other hand, the mature business model required for technology promotion is still being formed. The current market is in the early stage of rapid development, and the commercialization landing model, such as clear incentive mechanism, benefit distribution mechanism and common platform fee mechanism, has not yet been formed, which is difficult to support the large-scale promotion of the technology.

Take “blockchain + real estate registration” industry application as an example, at present, it mainly focuses on information registration to achieve interoperability at the information level, but does not realize the flow of assets and values.

“The social benefits and demonstration value are huge, but the profitability is relatively limited from the perspective of the current business model.” Yu Jianing, rotating chairman of the Blockchain Special Committee of the China Communications Industry Association, analyzed to Chain New that one of the constraints is that there is a certain degree of contradiction in information sharing and privacy protection.

“Real estate registration involves the participation of multiple parties, and information sharing in the blockchain multiparty ledger is needed in the process of multiparty participation, but how to protect the data privacy of each participating party after sharing? How to use technologies such as privacy computing to protect privacy security under the premise of information sharing and realize data available not visible is also a problem we need to solve in the future.” Yu Jianing said to Chain New.

In addition, the lack of relevant laws, standards and rules, data auditing, insurance and other supporting industries are still incomplete, which is also an important reason to restrict the development of privacy computing.

It is worth noting that there is no real killer application for privacy computing yet, and it is far from the time to distinguish the winner from the loser, thus the development potential is huge. With the current rapid development of the big data industry, efficiency, performance, cost and other comprehensive capabilities will be an important grip for all kinds of subjects to compete in the privacy computing industry.

Posted by:CoinYuppie,Reprinted with attribution to:
Coinyuppie is an open information publishing platform, all information provided is not related to the views and positions of coinyuppie, and does not constitute any investment and financial advice. Users are expected to carefully screen and prevent risks.

Like (0)
Donate Buy me a coffee Buy me a coffee
Previous 2021-07-01 02:02
Next 2021-07-01 02:44

Related articles