Is Data Special, and Are Data Assets Equally Unique?

In contemporary society, the topic of marketizing data elements is gaining increasing attention. As a new form of production factor, data isnot only required to circulate within enterprises but also aspires to flow throughout society to unleash its immense value. However, due to characteristics such as liquidity, shareability, risk potential, and complex ownership allocation, data struggles to fit into conventional management systems and regulations.

Do data assets, derived from data, also possess uniqueness similar to data compared to traditional assets?

For a complete review, please visit the [Data Trading Network] video account.

How does the uniqueness of data manifest?

Mr. Ma Huan: From multiple dimensions, data exhibits characteristics distinct from traditional assets, including replicability, consumption during use, physical space occupancy, and the generation of value through processing.

Firstly, data can be easily and inexpensively replicated without reducing or depleting its original content. Secondly, unlike material assets, data does not degrade or get consumed during usage; it can be reused repeatedly without losing value. Furthermore, data physically occupiesminimal space. While raw data lacks direct economic value, when large volumes are aggregated and transformed into information or knowledge through analysis and processing, significant value is generated, reflecting the potential inherent in data. These are manifestations ofdata's unique qualities.

Are data assets also unique?

Mr. Ma Huan: I'd like to clarify upfront that data assets do not inherently possess uniqueness; this assertion requires examination from the perspectives of data rights confirmation and data valuation.

Starting with data rights confirmation, once the ownership or possession of data is clarified, the allocation of benefit rights aligns with traditional assets. Users authorize companies to collect and utilize their data; the profits generated from products or services developed based on this data belong to the company, with no direct benefit to users.

Take Taobao as an example: users consent to Taobao collecting their personal and transaction data, which the platform uses to develop thebig data product "Business Advisor," providing valuable insights to merchants who pay for the service. The fee, however, isn't shared directly with users because they have already consented to the data collection and usage. Here, the ownership and usage rights of data are clearly separated, akin to traditional physical assets or intellectual property.

Considering clinical trials for new drugs, pharmaceutical companies pay substantial sums for participants' data. While this appears as a datatransaction, fundamentally, the company purchases physiological response data, a commercial act unrelated to the special nature of personal data. Participants earn high fees due to the risks they undertake in testing the drugs, not because their data possesses unique traits.

Thus, from the standpoint of data rights confirmation, data assets show no unique characteristics. Once ownership is established, the trading and management of data assets resemble those of traditional assets. Firms must abide by user authorization and relevant laws when utilizing data assets, ensuring their legitimacy and security. Moreover, certain conditions attached to transactions involving both tangible assetsand data assets can introduce "uniqueness," such as restrictions on asset disposal despite ownership transfer. Caution is urged when handling data.

Secondly, regarding data valuation, assets and resources hold different meanings for businesses. Human resources are not considered assets on the balance sheet, nor are office supplies like notebooks and pens recognized as resources. Amid the recent discourse on data entering the balance sheet, it's highlighted that only data with exploitable potential, reasonable reserves, and potential for economic benefits can be deemed a valuable resource, a premise underlying our recognition of data as a resource.

Although this viewpoint is widely accepted, many enterprises currently lag in data management capabilities, making their data reserves difficult to assess. Furthermore, determining the value of data poses a challenge. To illustrate using a physical commodity analogy, a hammer may hold great value in specific survival scenarios, yet its price remains grounded, tied to production costs within a manageable range. In contrast, data can hold varying values for different buyers under diverse circumstances. However, there is considerable variance in its practical applications, and the ultimate outcome may not necessarily yield tangible economic benefits. Therefore, attempting to uniformly value data based on its potential future benefits is impractical. The prevailing best method for valuation still leans towards cost-based approaches, indicating that data assets lack uniqueness compared to traditional assets in terms of both data rights confirmation and valuation.

In summary, data assets exhibit no particular uniqueness in terms of data rights confirmation or valuation.

From a Legal Perspective, How is the Uniqueness of Data Viewed?

Attorney Xie Hongtao: I fully concur with Teacher Ma Huan's perspective, and from my standpoint as a legal professional, I would like to add some complementary thoughts. The distinctiveness of data assets lies in their technical attributes or natural properties, yet in terms of their social attributes, they are not unique. Conversely, data elements and data assets focus precisely on the social attributes of data, rather than its technical characteristics. In contemporary society, amidst the advancement of digitization and informatization, the social aspect of data is receiving increasing attention. Data is regarded as a vital social asset or societal element, whose value is not only rooted in its technical processing and analysis but more so in its application and value creation across various domains of society.

Before delving into the topic, let me clarify the legal backdrop. Law, differing from morals, religions, and customs, is a social norm enforcedby state coercion, implemented by judicial and law enforcement agencies. Morals rely on societal pressure, while local conventions and customsare informally agreed upon practices, like lighting red candles during ancestral worship during Chinese New Year. This coercion is the fundamental distinction between law and other social norms – transgressing the law invites state-sanctioned punishment, enforcing compliance.

Secondly, laws stipulate general patterns of human behavior, abstracting and refining social life. The Civil Code categorizes various civil actsinto seven sections: property rights, contracts, personality rights, marriage and family, inheritance, torts, and general provisions, representing highly abstract, rationalized wisdom with strong stability and universality. This forms the legal foundation for property rights protection, mandating that property rights fall under state protection, established, utilized, and safeguarded according to legal norms, not arbitrarily created.

However, doubts surround the 'Data Twenty Articles' as the basis for data property rights protection. It is a policy document, not reaching the level of administrative regulations, unsuitable as direct evidence in litigation or claiming rights, functioning more as a component of public policy under “public order and good morals.”

Moreover, the NDRC's response to questions mentioned the 'Data Twenty Articles' aims to solve practical issues faced by market entities, innovating data property concepts by deemphasizing ownership and emphasizing usage rights, focusing on the circulation of data usage rights, and creatively proposing a framework for a data property rights system with "three rights separation" (data resource holding rights, data processing usage rights, and data product operation rights), constructing a Chinese characteristic data property rights system. Yet, this "creativity" doesn't align with existing legal frameworks, which already provide solutions for data property or ownership issues. Additionally, these "three rights" are vaguely defined, prone to confusion in practice, as "data resources" collected and "data outcomes" processed often become "data products," with overlapping definitions and low differentiation. The civil code's "possession, use, profit, and disposition" rights offer clearer expressions applicable directly.

Rooted in this legal foundation, all our property rights emerge. It is crucial to remember this as we delve into the historical evolution of property rights. Starting with physical property, we later saw the emergence of virtual or "artificially constructed" property, epitomized by intellectual property. Articles 859 of the Civil Code and Article 19 of the Copyright Law outline the principle of IP attribution—essentially, "who creates, enjoys." Leonardo da Vinci's creation of the Mona Lisa typically assigns him the IP, not Mona Lisa, unless otherwise agreed. The IP wouldn't naturally belong to Mona Lisa just because her likeness is depicted, but to da Vinci, the creator, following traditional IP systems. Similarly, in the age of automation and AI, while creative means may change, IP ownership adheres to "who creates, enjoys."

Data element property rights follow suit. The property of data outcomes or processed data naturally belongs to the entity that collects, processes, and forms it. However, it's crucial to distinguish details: the processed results differ fundamentally from the original information carriers or providers. Thus, providers or individuals don't naturally become proprietors of datasets, databases, or data outcomes, unless agreedotherwise.

Many misunderstand data assets, conflating data resource providers or original information suppliers with data asset owners. There's unnecessary hesitation in clarifying or asserting rights over data outcomes. It's important to recognize that while data resources may come from various sources, the subsequent processed output adheres to "who creates, enjoys," which is clear. Protection for data resource providers or individual privacy should be approached from the perspective of personal rights or dignity, not as property.

In conclusion, whether legally or societally, data assets do not possess uniqueness. Notably, the implementation of data "three rights" and data asset capitalization must align with existing legal systems and socioeconomic conditions. If they cannot be included within current legal protections or enforced by state coercion, they will prove unfeasible and unsustainable.

What is the relationship among data property rights, data resource holding rights, and data ownership?

Attorney Xie Hongtao: First, let's clarify a basic concept. The data we discuss, after collection and processing, is detached from its original information carrier or provider. Given today's discussion, I assert confidently that data property rights equate to data ownership. Rights to holddata resources, along with the data processing usage rights and data product operation rights mentioned in the 'Data Twenty Articles,' correspond to the "possession, use, profit, and disposition" rights within the civil code's property rights, backed by existing legal foundations. We need not shy away from or deviate from the civil legal system to "innovate"; we can directly apply civil law to reach this conclusion.

Mr. Ma Huan: I largely concur with Attorney Xie's view. Society has grown increasingly complex, leading to the segmentation of rights. In ancient times, with simpler social structures and relationships, the concept of ownership was singular – owning an asset meant complete possession. However, as society evolved and became more intricate, our understanding of rights deepened, dividing into different categories like usage rights, profit rights, and disposal rights. Present-day discussions on data ownership represent just a fraction of broader data rights.

Does data possess uniqueness, and does identical data exist in the world?

Mr. Ma Huan: This is an excellent topic that delves into philosophical realms. From a rigorous standpoint, I argue that there is no identical data in this world.

Many are familiar with the tale of Da Vinci painting eggs, illustrating that no two eggs are alike, much like how fingerprints or irises differ for each person. One could argue that hands generally resemble each other, all having five fingers, yet the fingerprints on each finger are unique. This suggests that there are no completely identical things in the world; even seemingly identical objects harbor subtle differences. Thisprinciple applies to data as well – every piece of data has its unique origin, storage location, and existence, rendering each one inherently unique.

In fact, when comparing two pieces of data, the mere act denotes they are two distinct entities. The uniqueness of data is an objective truth, absolute in nature.

Of course, we must acknowledge that commonalities can be derived from data under certain dimensions, but these are subjective and relative.

Regarding the current hot topic of incorporating data assets into balance sheets, are there potential risks involved?

Attorney Xie Hongtao: In recent years, factors such as the pandemic have led to an economic downturn, prompting a keen desire to leverage the digital economy to fill gaps left by the slowdown in the real estate sector. However, amidst this push, it's crucial not to blindly follow trends.The value realization of data elements is still being explored, and we must avoid repeating the mistakes of the past seen in the real estate market bubble. Currently, services related to rights confirmation, compliance, and valuation are widespread, but their efficacy, authenticity, and credibility must be meticulously scrutinized. Overvaluation or infringement of rights, once occurred, can significantly harm the interests of investors or enterprises.