China: Analysis and Commentary on the Key Points of the Judicial Interpretation (III) on Patent Infringement (Draft for Comments)
Published 25 December 2025
Sarah Xuan
On December 20, 2025, the Supreme People’s Court released the Interpretation (III) of the Supreme People’s Court on Several Issues Concerning the Application of Law in the Trial of Patent Infringement Dispute Cases (Draft for Comments) (hereinafter the “Draft”). The Draft intends to provide unified standards for key issues such as jurisdiction, standing, claim construction and infringement comparison, non-infringement defenses, regulation of bad faith litigation, and the coordination between damages and enforcement. Compared with the existing Interpretation (I) and Interpretation (II), this Draft is not a repetition of existing rules. Instead, it focuses on high-frequency scenarios such as e-commerce transactions, fluctuations in the stability of rights, evidentiary constraints, and litigation tactics, further codifying and categorizing areas that have long relied on judicial experience and case-by-case discretion.
This article will outline and analyze the key points of the Draft.
I. Standardization of Jurisdictional Rules The allocation of jurisdiction directly affects the litigation costs of the parties and judicial expectations. In response to the ubiquitous phenomenon of online shopping in practice, Articles 1 to 3 of the Draft establish a principle of prudent review. 1. Piercing Review of the “Substantial Connection” Principle: When a plaintiff attempts to lock in a forum by falsely listing co-defendants (such as joining unrelated distributors and manufacturers), the court shall conduct a substantive review of the “substantial connection” between the defendants and the matters in dispute. This rule aims to correct the imbalance of judicial resources caused by “forum shopping.” 2. Restrictive Interpretation of the “Place of Sale”: Addressing jurisdictional chaos in the context of e-commerce, the new rules explicitly exclude the “place of delivery chosen by the plaintiff” as a jurisdictional connecting factor. Previously, rights holders often established jurisdiction by placing orders for infringing products within specific judicial districts. The new rules restrict this to the “principal place of business” or “stable place of storage,” aiming to prevent rights holders from using the convenience of e-commerce flow to arbitrarily disturb the jurisdictional expectations of the accused party.
II. Coordination Between Procedure and Patent Validation Due to the “dual-track system” in China’s patent regime—where infringement trials and patent validation procedures are separate—dynamic changes in the stability of rights often lead to procedural volatility. Articles 4 to 7 of the Draft construct a tighter closed loop for this, specifically including: 1. Evaluation Reports as an Entry Threshold: For utility model and design patents that have not undergone substantive examination, the court may mandate that the plaintiff submit a Patent Evaluation Report. Failure to submit without justifiable reasons will result in the procedural consequence of “dismissal of the lawsuit.” This measure aims to filter “problematic patents” and prevent them from becoming tools for disrupting market order. 2. “Correction” Mechanism at the Second Instance: The Draft breaks through the previous time limit of “before the conclusion of oral arguments in the first instance,” allowing rights holders to change the asserted claims in the second instance based on effective patent validation results. This institutional design acknowledges the temporal misalignment between administrative validation and civil adjudication. By permitting changes in the second instance, it not only avoids misjudgments caused by poor procedural coordination but also achieves “substantive dispute resolution” at the appellate stage, reducing repetitive reviews triggered by changes in patent validity.
III. Reshaping the Boundaries of Claim Construction Claim construction is the core of patent infringement determination. The Draft enhances the objectivity of determining the scope of protection by introducing the concept of the “reverse doctrine of equivalents” and refining the standards for “estoppel.” 1. Technical Defect Exclusion Rule (Concrete Manifestation of the Reverse Doctrine of Equivalents): Article 10 of the Draft stipulates that if an accused technical solution falls within the literal scope of a patent, but in substance merely repeats a technical defect of the prior art that the patent specification explicitly claims to “overcome,” no infringement shall be found. For example, if a patent claims to improve a chemical reactor by solving the defect of “stirring shaft wobbling” in the prior art, and the defendant’s product precisely retains that specific “wobbling” structure without achieving the improved effect described in the patent, it should not be found to infringe even if it literally matches the patent features. This reflects the principle that “the scope of protection should be commensurate with the technical contribution.” 2. Strict Application of the Estoppel Principle: In balancing the interests of rights holders and the public, Article 11 adopts a presumption of “applicable unless explicitly rejected.” If a rights holder made restrictive statements to obtain a grant during patent prosecution or invalidation proceedings, such restricted portions shall not be reintegrated into the scope of protection in infringement litigation unless the statement was explicitly rejected by the examiner. This strengthens the duty of good faith during the patent application process.
IV. Optimization of Infringement Comparison and Evidence Acquisition To address persistent challenges such as the difficulty of “in-home evidence collection” for method patents and the “limited perspective” in design comparisons, Articles 14 to 18 of the Draft provide more operational measures. 1. Presumption of “Productization” of Method Patents: When an accused party solidifies the essential steps of a patented method into chips, algorithms, or product structures, and such solidification is indispensable for achieving the technical solution, the court may presume the implementation of the method patent. 2. Substitution of Physical Items with Technical Data: In cases where large equipment is difficult to move or seal, rights holders are permitted to use the accused party’s production process blueprints, operation manuals, and other materials as substitutes for physical items for comparison. This evolution of evidentiary rules balances the burden of proof for rights holders with the difficulty of inspection for the court.
V. Expansion of the Non-Infringement Defense System Articles 19 to 24 of the Draft conduct a refined restructuring of the prior art defense and the legitimate source defense, primarily in the following two aspects: 1. Constraints on Prior Art Combinations: The Draft explicitly prohibits defendants from using “hindsight mosaics” to forcibly combine multiple unrelated embodiments from a single document to assert a prior art defense. The defense must be based on technical correlations that a person skilled in the art can directly and clearly derive from said document. 2. The Cost-Sharing Tussle Under “Legitimate Source”: This is a focal point of the current solicitation. When a seller proves a legitimate source for a product, should they still be required to pay the rights holder’s “reasonable expenses” (such as attorney fees)?
Regarding this issue, Option 1 of the Draft leans toward “not supporting” such claims to protect the bona fide circulation chain. Option 2 leans toward “supporting such claims as appropriate,” aiming to prompt sellers to strengthen intellectual property compliance reviews in their supply chains by sharing the costs of enforcement. Option 2 considers the correction of cost misallocation in patent infringement governance; it posits that although a seller may be exempt from damages when a legitimate source defense is established, their transaction behavior objectively facilitates the circulation and diffusion of infringing products and increases the rights holder’s expenditures for evidence collection, investigation, issuance of warning letters, applications for preservation, or administrative complaints.
Therefore, Option 2 allows sellers to share reasonable expenses within a certain range to incentivize compliance measures such as improving supplier audits, retaining transaction vouchers, and setting up contractual indemnity clauses, making the supply chain more sensitive to infringement risks. However, the incentive effect of Option 2 depends on whether liability is commensurate with controllable behavior. If Option 2 is adopted, a more prudent path might be to introduce differentiated treatment centered on the “degree of due diligence” within the adjudication structure. For instance, the court could consider factors such as the traceability of the source, retention of vouchers, necessary verification of high-risk categories, profit and sales scale, and cooperation and rectification after receiving notice, thereby avoiding the transformation of uncontrollable risks into a burden for the seller.
VI. Restrictions on Bad Faith Litigation and Guarantee of Remedy Performance The Draft also provides corresponding regulations for the restriction of bad faith litigation and the performance guarantee of remedies: 1. Constitutive Elements of Bad Faith Litigation: Articles 25 and 26 codify “malicious filing of patent litigation” for the first time. In addition to stipulating liability for damages where litigation is filed with the clear knowledge of a lack of legal or factual basis for the purpose of obtaining illegitimate interests and causing harm, “timing factors” are specifically introduced as criteria for determining “bad faith”—such as filing suit while knowing the patent is invalid or has expired, or using litigation to exert pressure or cause delay during corporate financing, IPOs, mergers and acquisitions, or bidding. This reflects the judiciary’s close monitoring of malicious interference in capital markets. 2. Deterrent Effect of Late Performance Penalties: Regarding “non-enforcement of judgments” for injunctions, Article 29 introduces late performance penalties calculated on a daily or per-item basis. This increases the cost of non-compliant implementation of patents, transforming “cessation of infringement” from a paper judgment into a tangible market constraint.
Conclusion Overall, the Draft demonstrates a clear shift from “filling rule gaps” toward “refined governance.” First, it reduces the space for using procedures for strategic litigation layout through jurisdictional and procedural arrangements. Second, it further categorizes and standardizes rules regarding changes in validation results, evidentiary constraints, and non-infringement defenses to enhance the clarity of adjudication standards and the operability of remedy enforcement. Third, it provides institutional responses to order-maintenance issues in market operations through the design of rules on bad faith litigation and the bearing of enforcement costs.
For enterprises, adaptation can be prioritized in at least three areas: first, auditing online sales and warehousing chains and pre-setting jurisdictional and evidence-collection plans; second, establishing processes for evaluation reports and dynamic disclosure of validation status for utility models and designs to avoid procedural disadvantages; and third, forming evidence retention mechanisms (labels, contracts, purchase-sale-inventory records, promotional materials, platform pages, etc.) around prior art, legitimate sources, and manufacturing presumptions to adapt to potential changes in the evidentiary structure. Should the official version be finalized, its profound impact will lie in making litigation harder to use as a non-competitive tool while making genuine innovation protection more enforceable.
This article will outline and analyze the key points of the Draft.
I. Standardization of Jurisdictional Rules The allocation of jurisdiction directly affects the litigation costs of the parties and judicial expectations. In response to the ubiquitous phenomenon of online shopping in practice, Articles 1 to 3 of the Draft establish a principle of prudent review. 1. Piercing Review of the “Substantial Connection” Principle: When a plaintiff attempts to lock in a forum by falsely listing co-defendants (such as joining unrelated distributors and manufacturers), the court shall conduct a substantive review of the “substantial connection” between the defendants and the matters in dispute. This rule aims to correct the imbalance of judicial resources caused by “forum shopping.” 2. Restrictive Interpretation of the “Place of Sale”: Addressing jurisdictional chaos in the context of e-commerce, the new rules explicitly exclude the “place of delivery chosen by the plaintiff” as a jurisdictional connecting factor. Previously, rights holders often established jurisdiction by placing orders for infringing products within specific judicial districts. The new rules restrict this to the “principal place of business” or “stable place of storage,” aiming to prevent rights holders from using the convenience of e-commerce flow to arbitrarily disturb the jurisdictional expectations of the accused party.
II. Coordination Between Procedure and Patent Validation Due to the “dual-track system” in China’s patent regime—where infringement trials and patent validation procedures are separate—dynamic changes in the stability of rights often lead to procedural volatility. Articles 4 to 7 of the Draft construct a tighter closed loop for this, specifically including: 1. Evaluation Reports as an Entry Threshold: For utility model and design patents that have not undergone substantive examination, the court may mandate that the plaintiff submit a Patent Evaluation Report. Failure to submit without justifiable reasons will result in the procedural consequence of “dismissal of the lawsuit.” This measure aims to filter “problematic patents” and prevent them from becoming tools for disrupting market order. 2. “Correction” Mechanism at the Second Instance: The Draft breaks through the previous time limit of “before the conclusion of oral arguments in the first instance,” allowing rights holders to change the asserted claims in the second instance based on effective patent validation results. This institutional design acknowledges the temporal misalignment between administrative validation and civil adjudication. By permitting changes in the second instance, it not only avoids misjudgments caused by poor procedural coordination but also achieves “substantive dispute resolution” at the appellate stage, reducing repetitive reviews triggered by changes in patent validity.
III. Reshaping the Boundaries of Claim Construction Claim construction is the core of patent infringement determination. The Draft enhances the objectivity of determining the scope of protection by introducing the concept of the “reverse doctrine of equivalents” and refining the standards for “estoppel.” 1. Technical Defect Exclusion Rule (Concrete Manifestation of the Reverse Doctrine of Equivalents): Article 10 of the Draft stipulates that if an accused technical solution falls within the literal scope of a patent, but in substance merely repeats a technical defect of the prior art that the patent specification explicitly claims to “overcome,” no infringement shall be found. For example, if a patent claims to improve a chemical reactor by solving the defect of “stirring shaft wobbling” in the prior art, and the defendant’s product precisely retains that specific “wobbling” structure without achieving the improved effect described in the patent, it should not be found to infringe even if it literally matches the patent features. This reflects the principle that “the scope of protection should be commensurate with the technical contribution.” 2. Strict Application of the Estoppel Principle: In balancing the interests of rights holders and the public, Article 11 adopts a presumption of “applicable unless explicitly rejected.” If a rights holder made restrictive statements to obtain a grant during patent prosecution or invalidation proceedings, such restricted portions shall not be reintegrated into the scope of protection in infringement litigation unless the statement was explicitly rejected by the examiner. This strengthens the duty of good faith during the patent application process.
IV. Optimization of Infringement Comparison and Evidence Acquisition To address persistent challenges such as the difficulty of “in-home evidence collection” for method patents and the “limited perspective” in design comparisons, Articles 14 to 18 of the Draft provide more operational measures. 1. Presumption of “Productization” of Method Patents: When an accused party solidifies the essential steps of a patented method into chips, algorithms, or product structures, and such solidification is indispensable for achieving the technical solution, the court may presume the implementation of the method patent. 2. Substitution of Physical Items with Technical Data: In cases where large equipment is difficult to move or seal, rights holders are permitted to use the accused party’s production process blueprints, operation manuals, and other materials as substitutes for physical items for comparison. This evolution of evidentiary rules balances the burden of proof for rights holders with the difficulty of inspection for the court.
V. Expansion of the Non-Infringement Defense System Articles 19 to 24 of the Draft conduct a refined restructuring of the prior art defense and the legitimate source defense, primarily in the following two aspects: 1. Constraints on Prior Art Combinations: The Draft explicitly prohibits defendants from using “hindsight mosaics” to forcibly combine multiple unrelated embodiments from a single document to assert a prior art defense. The defense must be based on technical correlations that a person skilled in the art can directly and clearly derive from said document. 2. The Cost-Sharing Tussle Under “Legitimate Source”: This is a focal point of the current solicitation. When a seller proves a legitimate source for a product, should they still be required to pay the rights holder’s “reasonable expenses” (such as attorney fees)?
Regarding this issue, Option 1 of the Draft leans toward “not supporting” such claims to protect the bona fide circulation chain. Option 2 leans toward “supporting such claims as appropriate,” aiming to prompt sellers to strengthen intellectual property compliance reviews in their supply chains by sharing the costs of enforcement. Option 2 considers the correction of cost misallocation in patent infringement governance; it posits that although a seller may be exempt from damages when a legitimate source defense is established, their transaction behavior objectively facilitates the circulation and diffusion of infringing products and increases the rights holder’s expenditures for evidence collection, investigation, issuance of warning letters, applications for preservation, or administrative complaints.
Therefore, Option 2 allows sellers to share reasonable expenses within a certain range to incentivize compliance measures such as improving supplier audits, retaining transaction vouchers, and setting up contractual indemnity clauses, making the supply chain more sensitive to infringement risks. However, the incentive effect of Option 2 depends on whether liability is commensurate with controllable behavior. If Option 2 is adopted, a more prudent path might be to introduce differentiated treatment centered on the “degree of due diligence” within the adjudication structure. For instance, the court could consider factors such as the traceability of the source, retention of vouchers, necessary verification of high-risk categories, profit and sales scale, and cooperation and rectification after receiving notice, thereby avoiding the transformation of uncontrollable risks into a burden for the seller.
VI. Restrictions on Bad Faith Litigation and Guarantee of Remedy Performance The Draft also provides corresponding regulations for the restriction of bad faith litigation and the performance guarantee of remedies: 1. Constitutive Elements of Bad Faith Litigation: Articles 25 and 26 codify “malicious filing of patent litigation” for the first time. In addition to stipulating liability for damages where litigation is filed with the clear knowledge of a lack of legal or factual basis for the purpose of obtaining illegitimate interests and causing harm, “timing factors” are specifically introduced as criteria for determining “bad faith”—such as filing suit while knowing the patent is invalid or has expired, or using litigation to exert pressure or cause delay during corporate financing, IPOs, mergers and acquisitions, or bidding. This reflects the judiciary’s close monitoring of malicious interference in capital markets. 2. Deterrent Effect of Late Performance Penalties: Regarding “non-enforcement of judgments” for injunctions, Article 29 introduces late performance penalties calculated on a daily or per-item basis. This increases the cost of non-compliant implementation of patents, transforming “cessation of infringement” from a paper judgment into a tangible market constraint.
Conclusion Overall, the Draft demonstrates a clear shift from “filling rule gaps” toward “refined governance.” First, it reduces the space for using procedures for strategic litigation layout through jurisdictional and procedural arrangements. Second, it further categorizes and standardizes rules regarding changes in validation results, evidentiary constraints, and non-infringement defenses to enhance the clarity of adjudication standards and the operability of remedy enforcement. Third, it provides institutional responses to order-maintenance issues in market operations through the design of rules on bad faith litigation and the bearing of enforcement costs.
For enterprises, adaptation can be prioritized in at least three areas: first, auditing online sales and warehousing chains and pre-setting jurisdictional and evidence-collection plans; second, establishing processes for evaluation reports and dynamic disclosure of validation status for utility models and designs to avoid procedural disadvantages; and third, forming evidence retention mechanisms (labels, contracts, purchase-sale-inventory records, promotional materials, platform pages, etc.) around prior art, legitimate sources, and manufacturing presumptions to adapt to potential changes in the evidentiary structure. Should the official version be finalized, its profound impact will lie in making litigation harder to use as a non-competitive tool while making genuine innovation protection more enforceable.