Summary of Aug 2025 Patent Invalidation Outcomes

The newsletter mentions that the CNIPA issued 633 patent invalidation proceedings in August 2025, and provides a table with the names of several patent holders and the decisions made in their cases. These companies include YKK Corporation, Milwaukee Heavy Industries, Daikin, Pfizer and others. It also highlights two specific PCT utility model cases, one of the case survive when they go beyond numbers and focus on clear structural solutions.

Looking at July and August together, patent invalidation cases in China show a clear trend: most patents that are challenged end up being completely invalidated. In July, there were 599 decisions, with about 62% fully invalidated, 15% partially invalidated, and only 23% upheld. In August, the number of cases rose to 633, and the results were even tougher: about 68% were fully invalidated, only 9% partially invalidated, and 23% upheld.

When broken down by patent type, the picture becomes clearer. Utility model patents are the most at risk, with more than 80% either fully or partially invalidated in both months. Design patents also face high risk: in July about 60% were completely invalidated, and in August this jumped to almost 74%, making them the weakest category overall. Invention patents perform better. In July they showed an even split between fully invalidated (46%) and upheld (46%), while in August the rate of partial invalidations increased to nearly 28%, showing that courts are more likely to adjust them rather than completely invalid them. PCT national phase patents had mixed outcomes. Invention patents moved from more partial invalidations in July (40%) to more full invalidations in August (53%).

Overall, the data shows that designs and utility models are the hardest to defend, while inventions are more stable but still face risk. For companies, this means better drafting and careful prior art checks are essential if they want to reduce the high chance of losing their patents entirely in invalidation proceedings.

Case Study: Two Paths of Utility Models

Why One Failed and One Survived

Utility models are popular in China for their fast grant and cost efficiency. But statistics show they also face the highest invalidation rate, which is often above 60%. Two recent cases illustrate why some collapse entirely while others can survive the challenge.

Case 1: 201990001262.6 – Impact Tool (Milwaukee Electric Tools)

• Outcome: Entirely invalidated.
  
• Reasoning: Claims (on camshaft size, hammer stroke length, and shaft weight) were considered routine parameter adjustments with no unexpected technical effect. Prior art already disclosed similar structures; increasing torque by enlarging stroke/shaft diameter was obvious. Patent specification did not show surprising effect from chosen ranges (20.75 mm, ≥25 mm, 0.5–1.0 kg).
  
• Key Lesson: Without specific technical effect or clear teaching-away from prior art, numeric optimizations in utility models are easy to invalidate.
  

Case 2: 201790000304.5 – Wireless Audio Transmission System (Shenzhen Furong Tech)

• Outcome: Partially invalidated – claims 1–2, 5–8 fell earlier, but claims 3–4 and dependent claims 9–16 (built on 3–4) were upheld.
  
• Reasoning: Prior decision (49150) had already found claims 3–4 creative because connector/slot features were not disclosed in cited art and not considered “common knowledge.” Later invalidation request against 9–16 failed when they depended on 3–4. Only when dependent on claims 1–2, 5–8 did 9–16 fall, since those base claims lacked inventiveness.
  

Key Lesson: A carefully drafted narrower core claim (3–4) saved the patent. Even in a utility model, structural details of connectors provided novelty not trivial to combine.

Comparative Takeaways

1. Numeric ranges vs. structural features: Case 1 (impact tool) relied on size/weight ranges → seen as optimization, invalidated. Case 2 (audio system) relied on connector structure → upheld, as not obvious.
   
2. Specification support matters: Case 1 failed to link claimed ranges with unexpected technical effects. Case 2’s connector features had clear structural benefits (secure fit, easy detachment) that prior art didn’t anticipate.
   
3. Partial survival strategy: Utility models often face high invalidation rates (60–70%). Survival is possible if at least one claim introduces non-trivial technical structure that resists “obvious combination” attacks.
   

Suggestions for Leveraging Utility Models

1. Draft for structural specificity： Focus claims on mechanical configurations, interlocking structures, connector mechanisms, or assembly methods that are not easily derivable from prior art. Avoid over-reliance on numerical ranges unless backed by experimental data showing a new effect.
   
2. Build layered claims： Use broad claims for coverage but include narrow dependent claims with structural/functional details. If broad claims fall, narrower claims may survive (as in Case 2).
   
3. Strengthen descriptions： Utility models often fail because the spec doesn’t explain why a parameter or feature matters. Always link the feature to a specific technical effect (e.g., reduced vibration, enhanced torque stability, secure fitting).
   
4. Use utility models tactically： Quick filing (fast grant, no substantive exam) is ideal for short-cycle industries (consumer electronics, tools, connectors). Combine with design patents and invention patents for layered protection.
   

In short: Utility models are fragile, but when drafted with clear structural distinctions and linked effects, they can survive invalidation attacks and serve as fast, cost-effective protection.

Closing Note

Thank you for reading this update. Please note that our figures may slightly differ from official statistics due to data processing methods. We hope the data driven insights into recent patent invalidation cases help you stay informed about emerging trends in IP enforcement and legal reasoning in China. If you have any suggestions or would like to discuss a specific case in more detail, feel free to reach out, we’re always happy to connect.

*This newsletter represents the author's independent data analysis and does not necessarily reflect the views of any affiliated law firm.