A $16 Trillion Global Asset Reset for Data IP — By Jili Chung

China’s “Data Intellectual Property” Experiment Points to New Openings in the AI Market

1. ParagrapThe Boundaries and Core Value of Data Intellectual Property (Data IP)

AI has accelerated the data economy into a new phase of growth. In that broader context, data is no longer merely cold code sitting on a server. It has become a core input to how global commerce is organized and scaled. Governments worldwide are now testing ways to turn sprawling data flows into assets that can be priced, traded and financed.

Within this race for commercial advantage, Data Intellectual Property (Data IP) remains an imprecise term. Yet as a new category of intangible asset in the AI era, its potential is difficult to ignore—and that is increasingly the underlying force driving policy agendas and legislative experimentation across jurisdictions.

Before debating where this goes next, one basic question comes first: What is Data IP?

A recent analysis by the frontier digital-rights research platform IP Dao draws a sharp line—both legally and economically—between “what Data IP is” and “what it is not.”

What Data IP is

Data IP refers to derivative data products created by enterprises or individuals who lawfully collect data and then apply specific algorithms and processing steps—such as cleansing, anonymization, desensitization, or rule-based integration—to produce outputs with commercial use cases and attributes of intellectual contribution. The claim is to exclusive rights over that derivative product.

Put differently, what the law aims to protect is the intellectual work embedded in the downstream processing, and the incremental value created through that work.

What Data IP is not

By contrast, raw data dumps, unprocessed user browsing traces, or data collected through methods that violate privacy or national security constraints do not constitute Data IP.

This distinction matters because the original policy intent is not to grant firms an absolute monopoly over raw data. It is to build a commercial incentive structure for lawful, compliant secondary development—so data can circulate under clearer rules rather than freeze inside corporate silos.

How Data IP Differs From Traditional IP

Traditional intellectual property—patents, trademarks and copyrights—typically protects relatively static inventions, marks or works. The boundaries are clearer, and exclusivity is often strong.

Data IP also shares familiar traits of intangible assets: it is non-physical, easy to replicate, and its value is often contextual. But its commercial value tends to emerge only through dynamic circulation among multiple actors and through continuous data interaction.

As a result, Data IP system design often leans toward a balancing act: protect intellectual contribution while maximizing efficient circulation, including more flexible allocation of usage permissions over time.

Why Data IP Is Viewed as a Large Commercial Prize

The core reason is structural: Data IP is the only legally credible foundation for turning data into assets—and ultimately financial instruments. Without clear rights boundaries, the “code” inside enterprise servers cannot realistically be converted into balance-sheet assets with defensible valuation.

Cross that threshold, and data can be priced, traded, pledged for financing, and used as legitimate fuel for training large AI models. That is the center of gravity behind the intense attention from both capital markets and the real economy.

2. Watching China: Early Practice in Data Assetization

To see how the opportunity could take shape in practice, it helps to look to Asia—particularly China’s recent push toward “data assets.”

After entering 2025, China began producing concrete indicators in its effort to build a data-factor market. The trajectory has shifted from academic debate toward scaled commercial experimentation.

Several data points illustrate the change:

• Steady market expansion: Official estimates and consulting calculations put the overall size of China’s data-factor market at roughly $27.8 billion in 2025—suggesting the market has achieved a baseline level of capacity.
• Organizational moves by major enterprises: By the end of 2025, nearly 80 large state-owned enterprises had set up dedicated data companies or specialized subsidiaries, aiming to refine and operationalize large datasets in verticals such as industry and transportation.
• Early signs of “data assets on the balance sheet”: 2025 was widely treated in Chinese industry as a pivotal year for bringing data resources “onto the books.” By the end of April, about 100 A-share listed companies had formally disclosed the capitalization of data resources in annual reports, with disclosed totals near $300 million.

Taken together, the signal is clear: data is starting to shed its prior identity as an internal IT cost center and is being reframed as a balance-sheet asset with recognizable financial attributes.

But for data to trade like a financial product, the market first needs a credible rights credential. That leads directly to the next core topic: China’s government-backed pilots for Data IP registration and rights confirmation.

3. China’s Data IP Pilots and Rights Confirmation in Practice

One reason China’s data-factor market has scaled so quickly is institutional: the rollout of official Data IP pilot programs.

If intangible, replicable data is to be valued, recorded in formal financial statements, and then circulated and traded legally, the first requirement is a clear mechanism for defining rights.

In practice, these steps follow a sequence: data moves from resource-ization, to assetization, and then toward capitalization.

After 2025, China’s approach expanded from localized testing into a nationwide push. Administratively, IP and data-governance authorities have coordinated across agencies to establish standards for official rights confirmation and registration.

• Geographic scope: Early pilot regions—such as Zhejiang, Beijing, Jiangsu and Guangdong—have accumulated operating experience in registration, pledges and financing.
• Volume: Official statistics indicate that by October 2025, cumulative issuance of Data IP registration certificates across pilot regions exceeded 40,000.
• Value: Based on these certificates, the total value of related financing enhancement and licensing transactions has approached $2.1 billion.

These certificates are not merely administrative records sitting on government shelves. In real transactions they function as commercial credentials. When firms seek to list products at data exchanges, or apply for secured financing from banks and other institutions, these registrations can become essential.

This top-down, pilot-driven model has opened a workable path toward data assetization and injected liquidity into an emerging market.

4. Global Legal Debate—and China’s “Three Rights Separation” Response

Speed, however, comes with friction. Rapid implementation has exposed the limits of existing legal frameworks—and triggered deeper debate in legal academia worldwide.

In Europe and the U.S., as in fast-moving Asian markets, the question persists: How should data rights be defined? The dilemma is familiar and stubborn: how to balance strong protection for data developers with maximum circulation and reuse.

Globally, most jurisdictions have taken a relatively conservative route—stretching existing legal tools through expansive interpretation:

• European Union: anchored in GDPR and the Data Act, with emphasis on compliant circulation and strict privacy protection.
• United States: more reliant on contract law, trade secret protection, and antitrust, with a bias toward market-based boundary setting.
• Many other jurisdictions: often lean on copyright or unfair competition doctrines to cover certain forms of derivative-data infringement.

These patchwork approaches may resolve near-term disputes, but they struggle to establish data as an independent asset class with stable, legible property rights.

That is why the question of whether Data IP should have standalone legislation has sparked unusually intense debate in China:

• Opponents argue that creating a separate right could distort the existing IP system, and that current tools—patent, trademark, copyright, trade secret and related laws—already provide an operating framework.
• Supporters counter that without a clear, independent rights credential, “data assetization” and “financialization” remain architecture without a foundation. Given the economic importance of data assets and Data IP inside AI ecosystems, they argue for clearer operating rules.

This debate is not ivory-tower theater. The intensity itself is evidence of the commercial stakes. As AI model training and new financial tracks search for lawful “data fuel,” the market potential—and the prospect of massive capital flows—forces legal systems to confront rights confirmation with urgency.

Against that backdrop, China has advanced a framework known as “three rights separation”:
​

• Right to hold data resources: confirming a party’s lawful control status over raw data.
• Right to process and use data: protecting the legitimate interests of processors who invest substantial computing and intellectual effort.
• Right to operate data products: granting developers the right to commercialize derivative data products and earn profits.

The framework de-emphasizes traditional ownership battles and shifts toward a pragmatic logic: who develops, who benefits. It offers a region-specific reference for jurisdictions grappling with how to expand circulation without collapsing incentives.

5. Looking Ahead: Trillion-Dollar Openings Across AI, Innovation and RWA

A global legal consensus is unlikely to arrive overnight. Capital, meanwhile, is already seeking breakthroughs on the technology side. When the lens widens beyond China’s policy experiments to broader global tech trends, one conclusion stands out: the maturity of data rights infrastructure will directly shape the trajectory of three frontier domains.

(1) AI models: compliant “fuel” and the economics of ROI

The most immediate impact falls on generative AI, which depends on absorbing and learning from massive, high-quality training datasets.

IDC projects that by 2029–2030, total global investment in AI could reach $1.2 trillion. Yet behind the headline numbers lies a persistent pain point: AI development faces high compute costs and steep data-access barriers.

Without clear rights infrastructure, firms face not only copyright and privacy risk, but also uncertainty over whether AI projects can deliver defensible ROI—pushing many efforts into commercialization dead ends.

(2) The Innovation Breakthrough: How Blockchain and IP Management Are Reshaping AI’s Commercial Value

Confronted with this industry bottleneck, and running parallel to China’s state-sponsored Data IP push, the global Web3 sector is actively incubating a new wave of highly disruptive foundational infrastructure.

Within this space, the most lucrative growth vector lies in innovative platforms that fuse blockchain technology with professional IP management workflows, offering turnkey solutions across diverse application scenarios.

Take AIPBridge (https://www.aipbridge.com/) as a prime example. The advent of these innovative protocol platforms strikes directly at a glaring structural void within the current AI ecosystem. They seamlessly integrate traditionally cumbersome and labyrinthine intellectual property management processes into the collaborative networks of AI development, effectively unlocking the commercial value of Data IP.

From a market perspective, this is a quintessential blue-ocean opportunity. As capital continues to flood the AI sector, the "red ocean" competition among developers of upper-layer AI technologies is destined to become increasingly cutthroat. In stark contrast, those who construct the robust bridges enabling the "compliant monetization and frictionless circulation" of data assets will ultimately command pricing power over the entire AI industry right from the data source.

To draw a parallel with the energy sector: offering this type of structural solution is akin to positioning oneself as an oil-producing nation within the AI economy.

(3) Web3 and Real-World Asset tokenization (RWA)

Beyond AI enablement, Data IP also energizes another trillion-dollar track: Real-World Asset tokenization (RWA).

Because data is natively digital, it is arguably an ideal asset substrate for RWA. Forecasts by institutions including BCG suggest that by 2030, tokenized assets could reach a global market size of $16 trillion.

With the right technical stack, enterprises can convert operational data generated by real industries—such as cross-border supply chains or highly sensitive medical R&D—into Data IP, then tokenize those rights and cash flows on-chain, allowing them to circulate in financial markets as RWA.

The result is not just “activating” dormant data on servers, but also offering investors a new asset class anchored in measurable real-economy performance—reshaping liquidity at a structural level.

6. The Next Wave of Intangible Assets

China’s recent progress in Data IP registration and in bringing data assets onto financial statements offers the global market a working case study—a model for rights confirmation and pricing.

At the same time, private-sector platforms combining blockchain and rights infrastructure provide a commercially viable path to implementation.

Data IP remains an evolving concept. But as a new class of intangible asset, its potential is broadly acknowledged. In a world confronting a possible $16 trillion market reset, technology giants investing in AI and multinational firms searching for digital transformation should move early to assess how this shift could rewrite their operating environment—and their balance sheets.