Search

本文回顧各國法院關於標準必要專利公平、合理、無歧視（FRAND）之權利金計算所作成之司法判決，從比較研究之觀點進行分析及歸納。自美國法院於2013 年作成第一個關於標準必要專利FRAND 權利金計算的判決，各 國法院在FRAND 權利金計算的議題上，已陸續發展出許多值得參考的判決與方法論。本文深入個案進行判決之研究，從中萃取重要之計算方法與論理，描繪出可供依循之FRAND 權利金計算架構，而各國法院採納之具體方 法論以「由上而下法」與「可比較授權法」為主流，有僅採用其中一種方法，亦有見兩種方法並用作為交叉檢驗者，決定出標準必要專利最終之FRAND 權利金費率。其後聚焦我國智慧財產法院105 年度民專上字第24 號 判決，並就飛利浦與國碩公司間損害賠償額之決定，以各國相關司法判決為借鏡，分別解析一、二審法院計算賠償數額之方法論。本文主張可比較授權之授權金應僅為計算合理權利金之起始點，仍應將系爭專利貢獻度納入考量進行分配，俾使個案中合理權利金之數額能適度，避免專利箝制與權利金堆疊的風險。故期待本文對FRAND 權利金計算之比較研究觀察與啟示，能作為資訊通訊產業界與我國司法實務界面臨相關爭議時之參考。 The article reviews judicial decisions of fair, reasonable, non-discriminatory (“FRAND”) royalties in various jurisdictions. Starting with Microsoft v. Motorola in 2013, the first FRAND royalty decision ever made, courts from different juris-dictions have addressed their opinions on how to determine FRAND royalties. After exploring rationales in each case, the article then extracts useful approaches and key implications for determining FRAND royalties in order to depict an applicable framework. Courts now primarily consider either “top down approach” or “comparable license analysis,” or even both as a reliable cross-check, to come up with final FRAND royalties for SEPs in suit. The article further focuses on Philips v. Gigastorage, a controversial patent infringement case where Taiwan Intellectual Property Court awarded the patentee (Philips) an outrageous amount of damages. The article respectively analyzes and discusses the methodology the court adopted in its first and second instance. The article then suggests that under the comparable license analysis, royalties from a comparable license should be considered as a starting point for the calculation, and that the court should further consider apportioning economic value of the infringed patent(s) from the entire patent portfolio. The consideration could help avoid risk of patent hold-up and royalty stacking. The implications and trends in the article may shed some light on future FRAND royalty calculation, for corporations in the global telecommunications arena as well as for the judiciary in Taiwan.

均等論（Doctrine of Equivalents）是決定專利侵權重要步驟之一，甚至被認為專利法中最重要的原則。實際上，由於其與專利公示作用的衝突，充滿不確定性，甚至被認為是專利法原則中，最困難且最不可預測的。當被控侵權技術超出申請專利範圍文義之外時，專利權人可依據均等論，對與專利均等之技術主張專利侵權。而判斷被控技術與專利均等的比對方式，目前我國司法實務，主要採取智慧財產局編訂之「專利侵害鑑定要點」，其中規範：若被控對象之構件，與對應申請專利範圍之要件，「係以實質相同的技術手段（Way），達成實質相同的功能（Function），而產生實質相同的結果（Result）時，則屬於該申請專利範圍要件之均等，兩者無實質差異。而所謂的實質相同，係指兩者之差異為該發明所屬技術領域中具有通常知識者所能輕易完成者。若比對之技術手段、功能、結果其中之一有實質不同，則不適用『均等論』。而此一比對手段、功能與結果的方式，一般稱為三步測試法」。然而，「專利侵害鑑定要點」對於如何適用三步測試法，並未有夠具體的說明。本文以智慧財產法院歷來判決作實證分析，並比對美國法，探討我國司法實務上對於三步測試法進一步的適用程序與規範，檢視其中的妥適性，並提出作者之觀察與建議。 The doctrine of equivalents is one of the most difficult and unpredictable doctrines in patent law to apply. By allowing patentees to gain exclusive right beyond the literal claims in their patents, this doctrine creates tension between the adequate protection of patent right and the public notice function in the patent system. Based on the Guideline for Patent Infringement Analysis proposed by Taiwan Intellectual Property Office, to determined patent infringement under the doctrine of equivalents, the Triple-Identity Test should be applied. Under this test, if an accused device performs substantially the same function in substantially the same way to obtain the same result as the asserted patent, the device infringes the patent. However, the information regarding the test is insufficient in the Guideline for Patent Infringement Analysis. It fails to explain how to apply the test, how to define the Way/Function/Result, and what the sequence of applying the three factors is. Without these detailed information, it will be desirable for the public to know how the Taiwan Intellectual Property Court apply the test. This article attempts to review the practice of the test by empirical study. Especially, this article surveys the patent infringement litigations in the Intellectual Property Court and tries to analyze the test implanted in the Court. Besides, this article also provides some comments and suggestions on the test in Taiwan.

本文針對專利侵權損害賠償制度之設計及實務上所採行的計算，進行實證研究，用以檢視制度之實施，並對相關現象提供解釋。 美國、台灣與中國大陸的法律與法院，在損害賠償計算的原理上相近，但在實務運作上有相當大之落差，其關鍵在於訴訟當事人對損害舉證的要求與法院採信的程度而定。美國專利訴訟判決自1980 年至2000 年之間，以所失利益（lost profit）為主要的計算方式（83%）。然而，自2000 年以降，採用合理權利金（reasonable royalty）之案件反而成為主流（65%）。台灣現行所採的「差額說」與「利益說」，因為舉證不易，司法實務上過份倚重證據來源與客觀性，賠償數額也有偏低的情形。中國大陸的判決則幾乎都是由法官依據法定範圍酌定其賠償額度。 本文之結論為：1.專利糾紛的損害賠償，在不同管轄範圍內，由於法院的經驗、專利權利範圍與市場影響的不確定，以及產業經濟與企業財務的基礎環境差異，而有不同的判決。美國法院的經驗與財經基礎環境都較為成熟，判決較具說服力；台灣與中國大陸的法院對於當事人之舉證，通常認為未能達到足以確信的程度；2.負舉證責任的一方，對於證據被法院採信的機率難以確定，以及財務與經營資料作為證據之提出，有洩漏營業秘密之顧慮，因此縱然所失利益之計算可以期待較高之賠償金額，因而實務上傾向以市場中通常之交易情形，例如美國的合理權利金，或兩岸由法院酌定之賠償額；3.台灣與大陸以小額損害賠償為絕大多數，高額損害賠償非常少見，裁定的金額有偏低的現象；4.大陸對於發明專利損害賠償之判決數量極其有限，以國內廠商間之實用新型及設計專利侵害之案件為絕大多數。外國廠商的專利訴訟則以市場的佔有為主要考量，包括以行政查處及邊境保護防制侵權產品的內國與國際銷售，並非以取得法院酌定之賠償為其訴訟之目的。 This paper performs an empirical study on the regulatory frameworks and practical measures of the court decisions of patent damages among US, Taiwan and China. The basic principles employed in assessing patent damages seem similar among the above jurisdictions. Nevertheless, the outcomes of court decisions vary significantly due to the burden and standard of proofs of losses suffered by the patentees. Since 2000, the approach of reasonable royalty has overtaken the lost profit as the dominant measure of damage awards in the U.S. 65% of awarded damages were based on the measure of reasonable royalty, as contrast to 83% of lost profits during 1980 to 2000. It is also found that the difficulties of proving the amount of actual loss of patentees, as well as the cause-effect relationship directly caused by the infringing conducts, the damages decided by the courts of Taiwan and China are in general much lower than expected. In China, most of the compensations are determined with the statutory damages ranged from RMB 5,000 to 500,000, and more than 98% are lower than RMB 300,000. It is concluded that: (1) Court decisions indicate the differences of the legal infrastructures, the experience of judges, the enforceability of patent types and market structures among different jurisdictions. The U.S. courts are more inclined to recognize the evidences on a relatively broad and reasonable basis, nevertheless, the courts of Taiwan and China are considered less experienced in such decisions. (2) The party bearing the burden of proof finds the uncertainty of convincing evidence and also the risk of disclosing proprietary information, thus it results that the damages are primarily statutory provisions in China. Furthermore, Taiwan’s courts tend to utilize objective, though may not be proper, financial data as the base for damage measures. (3) Small damages are commonly granted in the courts of Taiwan and China. (4) In China, most of the patent lawsuits occurred among domestic companies with the new model and design patents, only very limited cases are associated with invention patents. So far as most international firms are concerned, instead of seeking monetary compensations, their litigations are mainly focused on acquiring the market shares by preventing the distribution of infringing products in China, and in the meantime, reducing the exportation of infringing products from China. It therefore leads to the frequent use of administrative enforcement, injunctions and boarder measures, in lieu of the damage claims in the judicial courts.

我國法學界向來重視抽象的原理原則，對於「先驗規則」（學者所提出之原理原則）的存在，甚少透過實際狀況加以驗證。法學知識的累積，固然可以來自於不同學說的論辯，然而如果缺乏充分與客觀的證據加以支持，學說理論的討論，就會變成各說各話的爭辯。再加上繼受自外國法律制度的影響，本土的法學論述與依據尚未建立，那麼台灣法學不僅無法在國際上與各國學說及實證相提並論，深耕本土法律的理想，更是難以達成。法律作為一種社會規範，其終極目的在於影響人類行為趨向公平正義，對於社會現象的觀察，自應根據特定社會與人類行為所發生的實際結果而為判斷，換言之，法律規定是否妥適，學說理論是否完備，只有經由實際證據與資料的蒐集及分析才能有效判斷，只有通過科學方法驗證的先驗學說或規則，才能確立其知識價值與社會接受性。 基於上述的體會與觀察，本文認為法學研究的發展方向，除了向來的法律邏輯演繹以及法律文字譯註之外，亦應嘗試跨領域的學科整合，進而探究法律的真實與擴展法學的研究，其中實證研究更是帶領法學走向嶄新領域不可或缺的觀念與研究方法。因此，本文說明法學實證研究的必要性及其內涵，亦提出法學實證研究的主要研究方法，以及如何建構法學實證研究的環境，亦即如何藉由法學教育的改革，培養法律人具備實證研究的思考與研究技巧，進而深耕本土法學議題，真正達到台灣法學國際化的目的，讓台灣法學建立自我的論述，進而與國際接軌，在本土生根。 Legal scholars are much more used to statue interpretation than empirical analysis of law in Taiwan. Although legal knowledge may be accumulated by counterarguments that are based on different doctrines, there is a lack of common platform for scholars to discuss the legal issues with objective data and evidences. Moreover, the legal theories and codes of Taiwan were primarily inherited from Germany and Japan in the last century, and the infrastructure and culture of our legal system which characterize the domestic judiciary are undergoing the changes gradually without distinguishing the features of this jurisdiction. As the law is never independent from the social constructs, the main purpose of the legal rules is to set up a social model of justice that guides the people to the rule of law. The consequences of legal regulations can be practically evaluated by collecting and analyzing the data or evidences from our observation and surveys. In other words, whether the legal regulations meet the needs of a society and serve the purposes of rule-making should be verified by empirical examination. The empirical legal study is one of the most eminent legal scholarships, which is not only employed as a research methodology but also a practical approach of discovering the realities of our society under the law. This article points out the difficulties and the connotations of empirical legal study and highlights this research approach. It contends that reformation of current legal education, redefinition of our local needs of empirical study, establishment of research centers, etc., will assist us to overcome the perplexity of empirical study. As a result, the legal experience gained from Taiwan can stand out for the review and knowledge sharing of international legal community.

本文所欲探討者，乃美國普通法所建立之「律師－委託人特權」 （attorney-client privilege）與「工作成果豁免權」（work product doctrine/immunity）等原則，適用於專利訴訟，以保護特定資訊免於被揭露之現況。第一部分將先分別介紹此二概念之基本內涵。第二部分則針對專利訴訟中，適用此二概念所引發之爭議與相關問題，予以分析與討論。在律師－委託人特權之適用方面，主要探討之主題包括：適用對象是否及於專利律師或專利代理人，與保護客體是否及於包含事實及技術之溝通。而在工作成果豁免權之適用方面，則討論其保護客體是否及於專利申請過程中，專利律師所製作之文件。最後，除說明對其棄權之情形外，亦討論當事人在放棄律師－委託人特權的同時，是否也等於放棄了工作成果豁免權的問題。 This article discusses the application of attorney-client privilege and work product doctrine/immunity to patent litigation to protect some information from being disclosed. First, the authors will introduce their basic concepts. Then the issues and related questions which result from their application to patent litigation will be discussed. The topics related to attorney-client privilege consist of whether it is applicable to patent attorneys or patent agents, and whether it protects factual and technological communications. The topic regarding work product doctrine/immunity is whether it protects documents produced by patent attorneys during patent prosecutions. At last, in addition to explain circumstances of waiver, the authors will discuss whether work product immunity would be waived as well if one waives attorney-client privilege.

當申請專利範圍與被控物兩者之間存在非實質性差異時，即成立均等構成專利侵害。美國最高法院於Graver Tank 一案中使用了我們所熟知的功能／方式／結果三部測試法來決定實質差異，此即我們所熟知的均等論。美國最高法院更於Warner Jenkinson 一案中再一次確認均等的決定應採逐項比對的方式。 逐項測試法將比對範圍放在構成要件上有其優點，然而，透過對美國聯邦巡迴上訴法院判決的研究，本文發現採逐項比對的方式有其不足之處，對侵權與否之認定甚至會造成截然不同的結果。本文擬從美國聯邦巡迴上訴法院的判決，經由系爭技術層面適用均等測試之探討，指出逐項測試法不盡周延之處，同時探討均等測試採整體觀察法的必要性。 To constitute patent infringement under doctrine of equivalents, it dictates that there must present insubstantial difference between claimed invention and accusation. In Graver Tank, U.S. Surpreme Court utilized the so-called functionway-result tripartite test to characterize the equivalency. There are two approachs before CAFC: one is by element-by-element test, the other is by invention as a whole. The outcomes of infringement analysis will hinge upon what accused counterpart should be the subject of comparision. Elemental approach reveals advantages when focusing counterpart on the component basis. However, we found that this approach shows incompletely and adversely result during analysis after reviewed the rationales underlying cases before CAFC. Via technical discussion to the application of doctrine of equivalents, this article shows the incomprehensive perspective of elemental approach and explores the entirety approach as a necessity under certain circumstance.