How to Outsmart the Trap of Chinese Support Requirements

In a previous article, the authors showed that Chinese courts ask a different question than U.S. courts — not whether the inventor possessed what is recited in the claim, but whether the scope of the claim matches the Technical Solution that the invention actually contributed to the advance of the art. This difference in underlying premise leads to one conclusion: a U.S.-style specification needs to be adjusted when filed in China. Below, the authors propose several strategies, offered as their personal views.

Drafting the Background of the Invention

In China, the background of the invention is not merely scene-setting. It is the evidentiary basis from which the court determines the Technical Problem the invention is understood to solve, and that Technical Problem in turn determines the scope of the Technical Solution the claims must recite. Two contrasting Supreme People’s Court decisions show what happens at each extreme.

(1) When the Background Is Drafted Too Specifically

The first is a case involving ELECON ASIA S.A.’s Chinese invention patent — “a carriage for the horizontal transfer of a vehicle in an automated mechanical parking facility” — widely known in the Chinese patent community as the Elecon case. In its 2014 retrial decision ((2014) Xing Ti Zi No. 13), the Supreme People’s Court reversed the Patent Reexamination Board’s Decision No. 14538 as well as the first- and second-instance judgments, and invalidated independent Claim 1 for violating the support requirement. The invalidated claim was extremely long — likely more than two pages if translated into English — yet this abundance of limitations in the independent claim did not save it.

The background of the invention in this case’s specification identified four deficiencies in the prior art — reliability, transfer speed, parking-space efficiency, and cost. The Supreme People’s Court’s reasoning was direct:

“Based on the specification, the patent at issue is required to simultaneously solve technical problems in four respects: reliability, transfer speed, space reduction, and cost reduction. Independent Claim 1 must recite all of the essential technical features that simultaneously solve the aforementioned technical problems. Decision No. 14538 (the Reexamination Board) and the first- and second-instance judgments, based on the technical features recited in independent Claim 1, found that the technical problem the patent at issue sought to solve was limited to improving transfer speed and reducing cost — this was an error in fact-finding and in the application of law.”

The mechanism at work here deserves close attention — it is arguably the most transferable lesson in this entire series. The court did not evaluate the Technical Problem separately from the background of the invention — it extracted the Technical Problem from the background. The moment the background identified four deficiencies, the court treated the invention’s contribution as necessarily spanning all four — regardless of how the applicant may have wished to frame things at the claim-drafting stage. The lower courts had allowed the patentee to argue that only two of the four (speed and cost) defined the invention’s contribution; the Supreme People’s Court held that this construction was impermissible — because the background of the invention itself had already fixed the Technical Problem along all four dimensions. Independent Claim 1, which recited a configuration addressing only some of those four problems, was necessarily missing essential technical features as to the rest, and was invalidated on that basis.

The lesson generalizes cleanly: in China, every deficiency mentioned in the background of the invention becomes part of the Technical Problem against which the Technical Solution of the later independent claim will be measured. The background of the invention is not the neutral scene-setting device it is often treated as in U.S. practice — it is closer to a self-imposed pledge as to claim scope.

(2) When the Background Is Drafted Too Sparsely

The opposite failure mode is less obvious, and at one point became fashionable for exactly the opposite reason — some practitioners drafted the background of the invention in only one or two sentences, on the theory that saying less leaves less room for it to work against you. This approach carries a distinct and often-overlooked cost — it can be fatal to an inventive-step (non-obviousness) defense.

China’s inventive-step analysis, like the EPO’s, generally follows a problem-solution framework — the technical problem the invention solves relative to the closest prior art is central to the analysis. If a patent’s own background of the invention and a competing piece of prior art describe essentially the same technical problem in similarly minimal terms, the problem-solution comparison collapses before it even begins — because there is no gap between what the two documents purport to solve.

This is close to what actually happened in the dispute we will call the Digitalize case here [China Reexamination Board Decision No. 35020]. In that case, Samsung Electronics succeeded in invalidating a Chinese standard-essential patent held by a Chinese company, relying on prior art that was the Chinese company’s own earlier patent. The invalidated patent and the prior-art patent shared the same background of the invention, which read as follows:

“Various different techniques exist for encoding and then decoding audio signals. However, improvements in performance, quality, and compression remain continuously desirable.”

Because the two documents identified the same general problem in the same generic language, the Reexamination Board — in commentary it later published — stated that such a brief background of the invention made it impossible to determine what specific technical problem the patent actually targeted, or what the claimed improvement was meant to achieve.

These two failure modes point to a single underlying principle: describing the technical problem in the background of the invention in too much detail invites a support-requirement violation, while describing it too sparsely invites an inventive-step violation under the problem-solution framework — because that framework requires distinguishing the invention’s problem from the problem addressed by the closest prior art.

(3) Toward an Optimal Background of the Invention

There is no universal formula, but the principle the authors have found most reliable is this: the problem stated in the background of the invention should correspond precisely — no more and no less — to the technical effect the independent claim is actually configured to achieve. If the independent claim’s essential technical features are directed at “improved speed,” the background of the invention should identify a technical problem relating to speed, and should not list a set of unrelated technical problems that the claim does not fully address.

Other technical effects the invention separately contributes — reliability, cost, and space efficiency, in the earlier example — can still be disclosed, but are better placed in the body of the specification than folded into the problem statement of the background of the invention. Otherwise, the scope the court treats as the fixed contribution of the invention becomes unnecessarily broad.

That said, applying this principle to an actual portfolio requires case-by-case judgment — particularly where an invention genuinely contributes to more than one technical problem, and the applicant must decide how to allocate that scope across a family of claims and a family of applications.

Preparing for Invalidation Proceedings — Reinforcing Claims and the Specification

(1) Why Recently Granted Patents Are at Elevated Risk of Invalidation for Support-Requirement Violations

Fifteen years ago, when the authors left for China, the support requirement was applied very strictly at the OA (Office Action) stage. This was, in fact, a large part of why the authors went to China in the first place. But at some point — the authors cannot pinpoint exactly when, but roughly five to ten years ago — Chinese examiners gradually began applying the support requirement more leniently during examination, and today’s OA-stage support-requirement review is considerably looser than it once was. The likely main cause, in the authors’ view, is that the China National Intellectual Property Administration (CNIPA) began encouraging examiners to focus their review on novelty and inventive step rather than the support requirement.

The problem is that this shift at CNIPA was never the product of a formal amendment to the Patent Law or the Examination Guidelines — it is simply an accumulated shift in practice over a long period — and Chinese tribunals and courts are under no obligation whatsoever to defer to that shift in examiner practice.

Moreover, the scope of amendment available in Chinese invalidation proceedings is extremely limited. So if a support issue that the examiner never raised becomes an issue at invalidation, and the tribunal or court adheres to the older, stricter standard, the patentee will find it very difficult to respond.

The result, as the SUNNY v. AAC case discussed earlier shows, is that Chinese tribunals have kept the old, strict support-requirement standard intact. In fact, in most of the decisions introduced earlier, the Supreme People’s Court of China has applied the support requirement more strictly than the lower courts and tribunals below it. In SUNNY v. AAC, if “the first lens” was found to violate the support requirement because it encompassed a “non-convex shape,” readers may wonder why the applicant simply didn’t amend to “a first lens having a convex shape.” In China, a claim amendment of that kind is not permitted unless another claim already recites that “the first lens has a convex shape.” In other words, the scope of permissible amendment is confined to what is already recited across the other claims. As a result, all eleven patents had their independent claims invalidated for violating the support requirement.

For reference, defining the scope of permissible amendment this narrowly is a distinctly Chinese feature of the system. In Korea and Japan, amendments based on what is disclosed in the specification — not merely in the other claims — are permitted.

(2) Reinforcing the Specification and the Claims

Given that amendment is difficult, and that support issues are a perennial feature of Chinese invalidation proceedings, the specification and claims need to be drafted with this reality in mind. It bears repeating: both SUNNY and AAC are Chinese companies. If the standard is applied this strictly even to Chinese companies, there is little reason to expect a more lenient standard for U.S. companies.

There are various ways to do this, and drawing on the lesson of the Amgen case is one worth considering. For example, one strategy is to recite narrower sub-concept technical features in separate dependent claims in advance, in preparation for invalidation proceedings. The important point is that this needs to extend beyond biotech and chemistry into electronics and mechanical engineering as well.

A Structural Deadline — The Constraint of Article 33

Every strategy discussed so far assumes there is still time left to make adjustments. But that time has often already run out, and the reason is worth stating explicitly — because it changes not only “how” this work should be done, but “when.”

Article 33 of the Chinese Patent Law prohibits any amendment that goes beyond the scope of disclosure of the original application. For an application entering China via the Patent Cooperation Treaty (PCT) route, that scope of disclosure is fixed to the international application itself — that is, to what was submitted as of the filing date. Once a PCT application enters the Chinese national phase, nothing that would be considered new matter can be added. Local counsel handling national-phase prosecution may be able to point out that an independent claim fails to satisfy the support requirement under Article 26(4), but by that point, there is usually no way left to fix it. By the time that stage is reached, the solution needed to already be in the document.

A few Supreme People’s Court decisions on new matter are worth noting: amending “semiconductor storage device” to “storage device” was held to constitute new matter ((2010) Zhi Xing Zi No. 53). Likewise, changing “circular bolt bore” to “circular hole” was also held to constitute new matter ((2013) Xing Ti Zi No. 21).

In the end, the time to technically tune a specification that was drafted to perfection for the USPTO, so that it can serve Asian jurisdictions as well, is not when the PCT application enters the national phase — it is when the PCT specification itself is being drafted.

Keep ‘Technical Solution’ in Mind From the Start

The techniques discussed above — a disciplined background of the invention, careful claim drafting, and the rest — are not independent tricks. They all follow from a single premise: what Chinese patent law protects is not evidence that the inventor once possessed an idea, but the Technical Solution. A specification drafted with this premise in mind — starting at the PCT drafting stage, before Article 33 closes the door, rather than at the office-action-response stage after it has already closed — will hold up far better whenever it is actually tested, whether in examination or in invalidation proceedings.

Image Source: Deposit Photos
Image ID:188290210
Copyright:Elnur_