Author | Kevin Han
Over the years, I have been accustomed to understanding intellectual property (IP) protection through the lens of legal norms and judicial practice—compliance boundaries, rights protection, infringement determination, liability allocation, and the like. These issues constitute the core of an IP lawyer's daily work and have shaped my fundamental understanding of the IP system.
Recently, I had the privilege of attending a two-day seminar led by the renowned economist Professor Xiong Bingyuan, where we discussed a number of classic legal cases, including the trolley problem, the O.J. Simpson case, and the Xu Ting case. Through these cases, Professor Xiong revealed that behind the law, there exists a deeper layer of economic logic.
Frankly, this analytical framework challenged the assumptions I had formed through years of legal education and practice. For instance, the function of law is not always to achieve fairness and justice, but rather to resolve conflicts of interest; honesty, virtue, and ethics are, in a sense, instrumental arrangements; and assigning different values to human life is, at times, a necessity. These cold economic propositions were admittedly unsettling at first. Yet upon further reflection, I had to concede that they carry considerable explanatory power when it comes to many real-world issues.
This prompted me to set aside my legal perspective and examine IP protection through the lens of economic logic. To my surprise, a number of long-standing and hotly debated issues became significantly clearer. Below, I share several reflections drawn from this learning experience.
I. The Legitimacy of IP Law Derives Not Only from Rights Protection, but More Fundamentally from Innovation Incentives
Traditional discourse on the IP system tends to emphasize its nature as a set of "rights"—patent rights, trademark rights, copyrights, and the like are civil rights that right holders enjoy under the law and that deserve legal protection. This characterization is, of course, correct. But if we stop here, we fall short of fully articulating the institutional value of IP law.
Whether we are talking about a technical solution, an established brand, or a literary or artistic work, the process of creation typically involves substantial investment—not only in capital, but also in time, intellectual effort, organizational resources, and the costs of trial and error. By contrast, imitators and copiers can often enter the market at far lower cost, rapidly profiting from others' achievements.
If the law were to be entirely permissive of such low-cost imitation, right holders would struggle to recoup their upfront investment, and rational market participants would gradually shift from "innovating" to "waiting, imitating, and copying." The consequence would not merely be harm to any single right holder, but a decline in society's overall supply of innovation, ultimately undermining technological progress, industrial upgrading, and wealth creation.
From an economic standpoint, therefore, the core function of the IP system lies in granting innovators exclusive benefits of a defined scope and duration, thereby enhancing the expected returns on innovative activity and incentivizing continued innovation. In other words, the legitimacy of IP protection rests not solely on its defense of vested rights, but on its service to a larger institutional objective: promoting innovation and advancing society as a whole.
Of course, the exclusivity inherent in IP inevitably imposes certain constraints on market competition. This means that stronger IP protection is not always better. Insufficient protection leads to inadequate innovation incentives; excessive protection may raise market barriers, impede the diffusion of technology, squeeze out legitimate competition, and even invite rights abuse and profit-driven litigation.
Seen in this light, what IP law truly pursues is not an abstract notion of "inviolable rights," but rather a dynamic equilibrium between innovation incentives and free competition, and between private rights and the public interest. Only within such a balance can the IP system claim both legitimacy and sustainability.
II. The Function of IP Adjudication Should Not Be Limited to Post-Hoc Remedies, but Should Also Provide Forward-Looking Behavioral Guidance
IP disputes are generally categorized as civil cases within the legal system, and damages have long been assessed primarily under the "make-whole" principle of civil law. While this is theoretically sound, understanding the judicial function in IP cases solely through the lens of traditional civil damages is, in practice, insufficient.
The reason lies in a highly characteristic challenge in IP infringement cases: the right holder's actual losses are often difficult to prove with precision, and the infringer's unlawful gains are equally hard to ascertain in full. As a result, in many cases, even where the right holder obtains a favorable judgment, the harm is not adequately compensated, and the cost of infringement for the wrongdoer is not high enough to serve as an effective deterrent.
In recent years, the state has continuously strengthened IP protection by raising damage awards and introducing punitive damages. However, in practice, factors such as the statutory caps on damages, the stringent conditions for applying punitive damages, and uncertainties at the enforcement stage mean that right holders still face considerable pressure in asserting their rights. For many companies, the real-world experience of IP enforcement is that litigation represents more of an expenditure and a cost than a source of meaningful market benefit.
From an economic perspective, IP cases differ from ordinary civil damage cases in an important respect: what they protect is not merely the private interest at stake in an individual case, but also the incentive to innovate in the future. It is precisely for this reason that the function of IP adjudication should not be confined to "resolving disputes," but should also assume the institutional responsibility of shaping market expectations and safeguarding the innovation ecosystem. Specifically, judicial decisions should send a series of clear signals to market participants: what kinds of innovation merit protection, what forms of imitation constitute infringement, what competition should be permitted, and what price infringers will pay. Once judicial outcomes form stable expectations, they serve not only to resolve past disputes but also to influence future market behavior. On this point, I am increasingly inclined to believe that IP adjudication needs a more pronounced "forward-looking" orientation.
Many judges exercise caution—even conservatism—in setting damage awards, and there are practical reasons behind this. A common concern is that excessively high awards could cause judicial remedies to be co-opted as profit-seeking tools by right holders. This concern is not without basis; instances of "profit-driven enforcement" have indeed occurred, and institutional design must guard against incentive imbalances. However, if we broaden our perspective from individual cases to the overall innovation ecosystem, it becomes clear that we should not let the fear of abuse hold back progress. Further strengthening IP protection and intensifying the crackdown on willful infringement is more consistent with the common good of humanity—social innovation and healthy development.
That said, excess in any direction is undesirable. IP protection must adhere to core principles: it must not impede the diffusion of technology, must not squeeze out legitimate competition, and must, on balance, be more conducive to incentivizing innovation and curbing opportunism. Under these principles, even if right holders obtain greater returns through enforcement, such robust judicial protection remains a sound institutional choice.
III. In the Age of Artificial Intelligence, Realizing the Economic Function of IP Law Faces New Challenges
The development of artificial intelligence (AI) has ushered in a period of intense new questions for the IP system. Whether it is the protectability of AI-generated content (AIGC), the boundaries of lawful use of training data, the allocation of platform liability in algorithmic recommendation and content distribution, or the attribution of liability for AI-related infringement, these issues have already exceeded the bounds of traditional IP legal analysis.
What makes these issues so complex is that they simultaneously touch upon the two most fundamental aspects of the IP system: how to maintain incentives for innovation, and how to recalibrate the boundaries of rights and the distribution of interests under new technological conditions.
(1) AIGC Tests the Boundary-Setting Capacity of Copyright Law
If the protectability of AIGC is categorically denied, it may dampen the willingness of users to create with the aid of new technologies. After all, in many real-world scenarios, AI is no longer merely an automated generation tool—it is a deeply integrated creative assistant, and in some cases, even an extension of the creator's expression. Yet if copyright is indiscriminately granted to all AIGC, it risks an excessive expansion of rights boundaries. Particularly in an era of ubiquitous AI and the mass production of low-threshold, high-frequency, bulk-generated content, overly broad rights could not only shrink the public domain but also significantly raise transaction and usage costs.
On this issue, therefore, a simplistic debate over "to protect or not to protect" is of little value. The more viable approach is to return to the foundational logic of copyright law and examine whether a particular user has invested a sufficient degree of intellectual labor in the creative process, and whether that labor is enough to support legal protection.
(2) The Training Data Question More Acutely Illustrates the Interest-Balancing Challenge of IP Law
The training data controversy more vividly demonstrates the tensions within the IP system in the age of AI.
If a strict authorization model is uniformly applied to training data, it may appear to favor the interests of original right holders. But in terms of industrial impact, this would raise market entry barriers, further concentrating data and computing resources in the hands of a few large platforms and capital holders, thereby reinforcing technological monopoly. Conversely, if the use of training data is entirely unregulated, the expected returns for original creators, content producers, and database investors will be undermined, which in the long run is equally detrimental to content supply and sustained innovation.
This is therefore not simply a question of "whom to protect and whom to restrict," but a classic problem of institutional efficiency: both overprotection and underprotection can lead to misallocation of resources. The key lies in designing a regulatory framework that better serves overall efficiency and long-term innovation.
In recent years, litigation and policy debates over generative AI training have proliferated across jurisdictions, and regardless of whether the conclusions favor technology developers or right holders, they invariably provoke significant controversy. This tells us at least one thing: the existing IP system has indeed encountered a clear inadequacy in addressing the new challenges posed by AI.
(3) In the AI Era, IP Protection Must Shift from "Mechanical Protection" to "Functional Calibration"
I do not have perfect answers to the questions above. But I am increasingly inclined to believe that in the age of AI, realizing the economic function of IP law can no longer rest on the mechanical protection of individual rights. This does not mean that individual rights are unimportant, nor that the legitimate interests of creators, inventors, and content producers can be readily sacrificed. On the contrary, their investment and returns should continue to command institutional respect and effective protection.
At the same time, however, we must recognize that AI has become a major force driving productivity, reshaping industrial structures, and enhancing national competitiveness. Against this macro backdrop, the question the IP system must answer is not merely whether a particular right should be protected, but what behaviors the system should encourage and what behaviors it should restrain. If a particular form of data use or technological application is primarily aimed at advancing technology and does not substantially substitute for the original work or significantly disrupt normal market returns, then perhaps the system should afford it a higher degree of tolerance. Conversely, where conduct involves willful substitution, outright appropriation, disruption of market order, or clear erosion of innovation returns, stricter regulation is warranted.
In this sense, what IP protection truly requires in the age of AI is neither simple strengthening nor simple weakening, but a recalibration grounded in institutional function. The ultimate objective of the IP system has never been to absolutize existing rights, but rather, through measured protection and reasonable limitations, to ensure that innovation continues to flourish and that new technologies can more effectively serve the overall welfare of society.
IV. Conclusion
Examining IP protection from an economic perspective, my most immediate takeaway is this: the legitimacy of IP protection is not weakened by such analysis—it is reinforced. The IP system matters not merely because it protects any particular right holder, but because it is integral to whether a society can sustain innovation and continue to create wealth.
In the past, I tended to think about IP issues primarily from the standpoint of laws, regulations, and judicial application. Now, I find myself habitually asking a further question: What behaviors does this institution or practice incentivize, what behaviors does it restrain, and what is its ultimate impact on the future innovation ecosystem?
Taking this a step further, such questioning is not only relevant to institutional design. It applies equally to individuals, families, businesses, industries, and indeed to society and the nation as a whole.
