中國China Patent OfficeCNIPACourt CasesInvalidationInventivenessPatentSupreme People's CourtTop 10 IP Case
2026年6月25日

No More Hindsight Bias: China’s Supreme People's Court Reins in Patent Inventiveness Assessments

For foreign applicants navigating patent litigation or invalidation in China, a common frustration is hindsight bias. Too often, CNIPA or lower courts strip down a patent into isolated technical features, find those features across disparate prior art references, and declare the invention to be "obvious". To counter such hindsight reasoning, the core criterion has always been whether a person skilled in the art would readily conceive of such technical solution.

In practice, that standard has not been applied consistently. Examiners and judges are often tempted to rely on their own intuitive standard, or worse-yet, directly use hindsight reasoning, undermining the consistency and predictability of the inventive-step analysis.

A relatively recent landmark decision by China’s Supreme People’s Court (SPC) (2024 Zhi Xing Zhong No. 7101 ”the 710 Case”) signals a welcome shift toward a more structured, objective, and predictable inventiveness assessment. In this case, the SPC provides a systematic explanation of how “technical teachings” should be identified. In particular, the SPC emphasized that the analysis should be conductedholistically by considering the relevant technical problem, application scenario, and technical effects. In doing so, the Court imposes important methodological constraints on inventive-step determinations.

The Facts: Same Flexible PCB but Different Uses

The central issue in the case was whether prior art reference “Evidence 14” —which disclosed the use of a flexible printed circuit board ("flexible PCB") in a pressure sensor—could provide a technical teaching for the distinguishing feature of the patent at issue, namely, using a flexible PCB in a vibration-and-shock composite sensor to isolate vibrations transmitted from the sensor housing to the interior of the sensor.

The conclusions diverged significantly depending on the stage of the proceedings:

Lower Instances: Both the CNIPA (invalidation stage) and the first-instance court ruled the invention to be obvious. Their reasoning was that flexible PCBs are well-known, so replacing one with another is an "obvious alternative".

SPC: The SPC reversed the decision, disagreeing that there existed a "technical teaching". The SPC emphasized that simply pointing out that a technical tool is "known" is not enough to prove that a person skilled in the art would naturally think of applying it to a completely different context.

Comparison of the Reasoning Adopted at Different Stages

To better understand the factors underlying these divergent conclusions, the key distinguishing technical features were analyzed as follows:

Stage

Core Conclusion

Reasoning

Methodology Issue

Invalidation Proceeding

Obvious

Flexible PCB already known → could be substituted

Analogy based on technical means

First Instance

Obvious

Motivation to improve + known technology

Motivation substituted for technical teaching

Second Instance (SPC)

Not Obvious

Technical problem, application scenario, and technical effect all differ

Return to the technical context

The SPC's conclusion was not based merely on the observation that both the patent and Evidence 14 employed a flexible PCB. Rather, the Court focused on the substantially different technical contexts in which the flexible PCB was used.

In Evidence 14, the flexible PCB serves as a connection medium for constructing a modular sensor architecture. The underlying technical problem is how to allow individual components to be positioned independently, rather than being fixed on a single rigid printed circuit board. By flexibly interconnecting components located at different positions, the flexible PCB enables the formation of a modular structure in which each module can be arranged at its optimal location.

By contrast, the patent addresses a fundamentally different problem arising in harsh operating environments such as rail transit and wind power generation. In these applications, vibration sensors are typically connected through large, rigid cables that must be secured by cable cleats. The substantial mass and rigidity of such cable assemblies transmit vibrations from the sensor housing directly to the internal resonator, leading to low-frequency resonance, frequency-response distortion, and degraded sensing performance. The technical problem addressed by the patent is therefore how to reduce resonance caused by the sensor's external structure and prevent vibration transmission from the housing to the internal sensing elements.

To solve this problem, the patent replaces conventional rigid cable connections with flexible PCB connections between sensor modules, so that vibration transmission from the outer housing to the interior can be reduced or isolated. As a result, low-frequency resonance interference is suppressed, and a flatter frequency response is restored.

Therefore, although both solutions employ a flexible PCB, the application scenarios, technical problems, and technical effects are materially different. Under such circumstances, the mere disclosure of a flexible PCB in Evidence 14 does not provide a technical teaching that would lead a person skilled in the art to adopt the claimed solution of the patent.

The Supreme Court’s Conclusion: Evidence 14 and the provided evidence of common knowledge did not disclose the distinguishing feature 1 of Flexible PCB, nor did they provide corresponding teaching. As a result, claim 1 as a whole was not disclosed by the prior art nor common knowledge, and thus possessed inventiveness.

Key Principles Established by the Supreme People's Court

  1. Prior Art Should be Considered “As a Whole”

Whether the prior art provides a technical teaching should be determined through a holistic assessment of the relevant technical problem, application scenario, and technical effects.

  1. Similar Technical Means Does not Equal a “Technical Teaching

Just because a prior art reference has similar technical means does not mean that a “technical teaching” exists. For example, generally there is no technical teaching if the prior-art feature’s function differs from the function of the claimed invention’s distinguishing feature.

  1. Don’t Presume "Readily Conceivable" Without Supporting Evidence

If a feature has not been disclosed in the prior art and is not shown to be common general knowledge, one cannot presume that the feature is readily conceivable by a person skilled in the art.

  1. Technical Problem Should be Narrowed to a Specific Application

Define the technical problem based on the specific application environment and the role performed by the relevant structure, rather than generalizing at an abstract level.

Does the Prior Art Provide a Technical Teaching? A 3-Part Test

Based on the Court's reasoning, we can derive following three-factor analytical framework:

Same Technical Problem: Does the prior art address the same technical problem as the patent in suit?

Same Application Scenario: Are the respective technical fields and application environments the same or at least consistent? ​

Same Technical Effect: Does the technical feature achieve the same technical effect in both technical solutions? ​

Only when all three factors are satisfied should a technical teaching be established.

Eagle IP Commentary

The 710 Case together with related earlier cases reveal that the inventive-step assessment is changing from a conclusory form of reasoning to a structured analytical approach.

First, from the perspective of judicial methodology, this case clearly distinguishes three different approaches which focus on different aspects:

  1. analogy approach: the existence of similar technical means;
  2. motivation approach: the possibility or incentive for improvement;
  3. technical teaching approach: the specific technical context.

Through this decision, the SPC effectively imposed institutional limits on the first two approaches and made clear that inventive-step determinations must be based on verifiable technical teachings. This serves to curb the hindsight bias that has frequently appeared in practice.

Second, at the level of analytical reasoning, the concept of a “technical teaching” is further broken down into several constraining elements: the technical problem, the application scenario, and the technical effect. These factors form an integrated evaluative framework. Only when all factors are satisfied should a technical teaching be established.

This structural approach effectively changes the inquiry of whether a person skilled in the art "would have readily conceived" the claimed invention from a subjective cognitive assessment into an analysis that must consider objective technical facts. This standard should significantly enhance the transparency and predictability of inventive-step determinations in China.

Third, the case also raises the Examiner’s burden of proof for establishing that a claimed feature would have been “readily conceived”. For example, if the prior art does not disclose the relevant distinguishing feature, and there is insufficient evidence that the feature is “common general knowledge”, then the Examiner should not conclude a lack of inventive step solely based on “experience” or abstract reasoning.

Accordingly, this case provides a clearer direction for future inventive-step analysis. Rather than focusing on the abstract question of whether a skilled person "could have thought of" the claimed solution, the inquiry should return to a more concrete question: does the existing prior art contain a genuine technical teaching leading to the claimed invention?

If you would like to have more information on this matter or would like to have our advice, please feel free to contact us at [email protected].

This article is for general informational purposes only and should not be considered legal advice or a legal opinion on a specific set of facts.

  1. 最高人民法院行政判决书:(2024)最高法知行终710号。 ↩︎

其他文章

Markush claims in China - what can be arbitrarily deleted during invalidation?

2018年12月4日
Since 2010, the China Patent Re-examination Board (PRB) has published the top 10 patent invalidation cases of the year in April of each year. The selection criteria are high social concern, significant impact on the related industry, or involve difficult legal issues and important examination criteria. Below is one of the top 10 cases that […]

AI is magical but NOT magic! Be specific in your AI patents!

2025年1月9日
What is the standard of sufficient disclosure for AI-related patents: A case study in China The rapid growth of AI patent applications present significant challenges to existing patent application and examination practices. One of them is satisfying the requirement for sufficient disclosure. The following case was featured in the 2023 CNIPA Compilation of Key Decisions […]

Twice the Trouble: Unraveling a Single Case of Dual Patent and Trademark Infringement

2024年1月9日
Can a rights holder sue the same infringer separately based on one single infringing act that infringes both trademark and patent rights? An interesting case this year from China’s Supreme People’s Court that addresses this specific issue ((2023)最高法知民终235号). Beijing Run De Hong Tu Technology Development Co., Ltd. (“Run De Hong Tu”) sued an individual named […]

Please “Mind the Traps”: Four Common Misconceptions About Hong Kong Patents

2026年4月13日
For many foreign companies interested in the broader Asian market, Hong Kong is a strategically important jurisdiction. However, because of its unique legal framework, domestic and foreign companies alike are confused by how IP protection works in Hong Kong. Missteps in such understanding can lead to critical, costly "gaps" in IP protection. To safeguard innovations […]

我們過去活動

Top crossarrow-right