The enforcement of standard-essential patents.
A possible limitation under the Charter.
THESIS - MASTER INFORMATION LAW
UNIVERSITY OF AMSTERDAM
JANUARY 2016
AUTHOR: K.G. van EKEREN
DATE: 4 JANUARY 2016
STUDENT NR.: 10791264
SUPERVISOR: S.J.R. BOSTYN
Abbreviations 3
Introduction
4
Chapter 1: Standardisation in practice 7
§ 1.1: Institutions 7
§ 1.2: Standards
8
§ 1.3: SEPs 10
§ 1.4: Institutions and SEPs: SSO IPR Policies 12
Chapter 2: Delineating Enforcement 17
§ 2.1: Enforcement of SEPs in the EU 17
§ 2.2: Antitrust scrutiny
19
§ 2.3: The role of SSO IPR policies
22
Chapter 3: The Charter, a thought experiment
29
§ 3.1: Limiting SEPs under article 16 of the Charter
29
§ 3.2: Counterarguments 34
§ 3.3: Practical implications of the Charter-discussion: Closing considerations
35
Conclusion 37
Bibliography
38
Books
38
Articles
38
Case-law
41
Legislation 43
Other 43
Webpages 45
Abbreviations
ANSI American National Standards institute CEN European committee for standardisation
CENELEC European committee for electrotechnical standardisation Dir. Directive
EC European Commission EP European Parliament EPC European Patent Convention
ESO European Standardisation Organisation
ETSI European Telecommunications Standards Institute FRAND Fair, reasonable and non-discriminatory
GPRS General Packet Radio Service
ICS International Classification for Standards IEEE Institute of Electrical and Electronics Engineers IETF Internet Engineering Task Force
IPR Intellectual property right
ISO International Organisation for Standardisation ISP Internet Service Provider
JEDEC Joint Electron Device Engineering Council LTE Long Term Evolution
NEC Nederlands Elektrotechnisch Comité NEN Normalisatie en Normen (name of the SSO) NSO National Standardisation Organisation OS Operating system
TFEU Treaty on the functioning of the European Union RFC Request for comments
Reg. Regulation
SDO Standards Development Organisation SIG Special Interest Group
SSO Standard setting organisation SEP Standard essential patent W3C World Wide Web Consortium
Introduction
Over the last decades, technological devices have become an increasingly normal part of our daily lives, and with that, the standards that govern the functioning of those devices. For something as simple as opening up 1
a webpage in a browser or establishing a connection between two mobile devices over the internet, a significant amount of protocols are needed that operate on standardised cables and hardware which are 2 3 4
built or configured in such a way that they are interoperable. However, it is well possible that there are patents covering technologies that are essential to these standards. This is where the term Standard-essential patent (SEP) comes from. SEPs are patents that exist in a context that is different from the ‘normal’ context for patents. To speak of them in one breath would negate the considerable differences that exist between them. Moreover, the underlying philosophies of standardisation and IPR-protection might be seen as conflicting. Standardisation intends to put ideas into the public domain, whereas protection of IPRs makes them private property. This view is clearly reflected in the older approaches taken by many SSOs to IPRs. 5
CEN en CENELEC published in 1992 that ‘any use of IPR by a standard is an anomaly, sometimes an
unavoidable one, which needs careful management’ . In 1932, when ANSI first addressed the issue of the 6
incorporation of patents in standards, they maintained that ‘as a general proposition, patented design or
methods not be incorporated in standards. However, each case should be considered on its own merits (…).’ SSOs have adopted ample patent policies in the meantime, which is testament to the fact that they can no
7
longer maintain this attitude towards IPRs. This moderation is reflected in the ETSI IPR policy where it states that it ‘seeks a balance between the needs of standardisation for public use (…) and the rights of the
owners of IPRs’’ and further evidenced by the fact that many IPR policies were developed between 1999 8
and 2002. This, however, does not mean that the conflicting nature of standards and patents has disappeared. 9
It merely means that SSOs, governments and other stakeholders have worked to create mechanisms that aim at reconciling standards and patents in practice. There exists a great diversity in opinions on the conflict between patents and standards. Kriegel argues for the complete elimination of formal standardisation
http://ec.europa.eu/competition/publications/cpb/2014/008_en.pdf
1
For a modest overview: http://www.rfc-editor.org/rfc-index.html
2
For an overview of, i.a., the UTP series: https://www.siemon.com/us/standards/13-02_over.asp
3
Such as the ethernet ports on a network interface card.
4
Falke (2000), p. 175
5
Ib id. CEN/CENELEC now has a patent policy due to the fact that an increasing number of
6
patent-based standards have been successfully developed. The SSO still pleads for a universal approach of standards, but meanwhile respecting the rights of patent holders. See: http://
www.cencenelec.eu/ipr/Patents/Pages/default.aspx EC Policy brief, p.2
7
ETSI IPR Policy, art. 3.1
8
Lemley (2002), p. 1904
(thereby omitting the problems with SEPs) in economies with strong network effects through the strict application of antitrust law. This view might be taken as one extreme end of the spectrum, the opposite of 10
which would be the possibility of ‘regular’ enforcement of SEPs without any additional considerations in statutory law or court practice for the distinct status of an SEP. Kahin proposes to address standards specifically as an exception to the power of the patent holder to extract damages and constrain potential infringers. The latter view is not far from the current legal reality. It is possible to restrict the enforcement 11
of SEPs through antitrust law. However, due to a growing public interest (and in some cases it might be argued, dependency), it is interesting to explore whether additional legal grounds for limitations can be found. This article explores whether the Charter can be directly invoked against the enforcement of SEPs. Specifically because the freedom to conduct a business has already been successfully asserted above the 12
right to IP and there is a considerable body of literature proposing the possibility of asserting human rights 13
over IPRs. 14
Because of the ample variety in standardisation processes, standards and the rules governing them, a separate chapter is devoted to providing an overview of standardisation in practice. In chapter 1, a categorical
exposition of SSOs, standards and SSO IPR policies is made. Also, the concept of an SEP is introduced by placing patents in the standardisation context. It serves to introduce and define standardisation and its constituents to the reader in order to be able to analyse the field from a legal perspective. Chapter 2 then zooms in on the legal framework for the enforcement of SEPs. The chapter starts out with the enforcement of SEPs by briefly specifying the possible actions and the restrictions built into statutory law. The second and third paragraph respectively treat the ways in which SSO IPR policies and antitrust law scrutinise these actions. Finally, attention is given to the way in which implementers that are confronted with SEP-based claims can defend themselves. The final chapter is presented in the form of a thought experiment in which article 16 (freedom to conduct a business) of the Charter is first interpreted as a possible limitation on injunctions for SEPs.The second paragraph then argues for the contrary. Finally, the practical implications of the interpretations in the first two paragraphs are discussed. Below is an overview of the questions posed by this article and the chapter/paragraph they relate to.
Main question
Can the enforcement of SEPs be limited under art. 16 of the Charter?
Kriegel (2006), p. 231
10
Kahin (2007), p. 7
11
laid down in art. 16 of the Charter
12
laid down in art. 17(2) of the Charter, Peers (2014), p. 455
13
E.g.: Brinkhof claims that patent law is subordinate to human rights law, when there is a conflict,
14
patent law must give way. (Brinkhof (2010), p. 153) Van Overwalle asserts that human rights potentially add to the public interest component of patent law and can therefore act as a limit to patent rights. (van Overwalle (2010), p. 241)
Subquestions
What is standardisation and what are SEPs? (Chapter 1)
How are SEPs enforced in the EU and how is this enforcement restricted by antitrust law? (Chapter 2)
Does the Charter provide a limitation on the enforcement of SEPs? (Chapter 3, §1)
What are the arguments against limiting SEP-enforcement under the Charter? (Chapter 3, §2)
What would be the implications if a limitation were to be endorsed in practice? (Chapter 3, §3)
The scope of this article is limited in two ways. First, it only deals with legal sources within Europe. 15
Second, it relates solely to compatibility standards in the computer networking and telecommunications industry. The latter limitation is incorporated for two reasons. First, because this is the industry where the most contentious IP issues arise. Second, because it is an industry with strong networking effects and 16
therefore highly (and ever increasingly) reliant on interfaces that connect systems of interdependent components. These interfaces are seldom standard-free and are the main drivers of information-access for 17
the public. Consequently, the interests of implementers of standards and the public are most easily affected by IPR conflicts in this area. Therefore, the proposition made in §3.1 and falsified in §3.2 is mainly 18 19
applicable to this industry. 20
Legal sources in the most broad sense, i.e. including bylaws, contracts and other non-public
15
sources.
Lemley (2002), p. 1903
16
Gawer & Henderson (2007), p. 1. These systems are also referred to as ‘platforms’.
17
See, for example, van Overwalle (2010), pp. 261 - 262
18
The proposition is not completely falsified but rather put into perspective by providing arguments
19
for the contrary.
It goes without saying that the assertion of SEPs can affect consumer and implementer interests
20
in any technological industry. However, the ultimate object we are dealing with in the ICT-industry is information and the transmission thereof. Both concepts that have acquired the character of a public good in the past decades and key to reducing information-asymmetry (e.g.: washing-machines and toasters are not fundamental rights and do not communicate information). Which is a good thing in the author’s opinion.
Chapter 1: Standardisation in practice
§ 1.1: Institutions
The EU divides SSOs into 1) formal, 2) quasi-formal and 3) consortia. The main characteristic of all categories is that they are groups of stakeholders, working together on a voluntary basis to produce standards. On a geographical scale, a distinction exists between national (e.g. NEC in the Netherlands), 21
regional (e.g. ETSI in Europe) and global (e.g. ISO) SSOs. Formal institutions, which will be referred to in 22
this article as SDOs (SSOs being the general denominator for all three categories mentioned above), are 23
SSOs recognised by governmental organisations or governments. They often have an area focus such as engineering or communications and can be united in a cooperative SSO where areas overlap. In Europe, a recognised SSO is a national or European SSO included in Annex I or II of Dir. 98/34/EC. While there 24
exists a great variety of SDOs, the European SDOs are CEN, CENELEC and ETSI. ETSI is an SDO for 25
technical specifications for telecom, broadcasting and information technology. They have developed standards such as the GSM, 3G and 4G specifications. ETSI was founded in 1988 as a private non-profit 26
association under French law. CENELEC is responsible for standardisation in the electrotechnical 27
engineering field. It grew out of a cooperation between national electrotechnical committees and was set up 28
as a de facto organisation in 1959. CEN is an overarching organisation that provides a platform for 29
cooperation in European standardisation (mainly the harmonisation of standards adopted by members of CEN) and works closely together with ISO as well. It was founded as a de facto organisation in 1961. The 30
cooperation between the three European SDOs is regulated in Reg. 1025/2012. The level of cooperation is 31
P&S Report, p. 31
21
ib id., p. 32
22
Derived from Biddle et al. (2012) and Updegrove, A.
23
Art. 1(7) and 1(8), Dir. 98/34/EC
24
Annex I of Dir. 98/34/EC
25 http://www.etsi.org/index.php/standards 26 Falke (2000), p. 10-11 27 http://www.cenelec.eu/aboutcenelec/whoweare/index.html 28 Falke (2000), p. 6 29 https://www.cen.eu/about/RoleEurope/ESS/Pages/default.aspx 30
Art. 1, Reg. 1025/2012. The regulation deals with a variety of topics relating to standardisation,
31
mostly determined by the technical overlapping of the respective work-areas of the SSO’s. The majority of 32
European NSO’s are incorporated into CEN, CENELEC and ETSI. NSO’s do not need a detailed 33
description here. For the purposes of this article, it suffices to say that European standards do not stand on their own. They need the adoption and implementation by member NSO’s in order to be effectuated in 34
member states. NSO’s are responsible for the coordination and supervision of standards on a national level. 35
Standardisation is considered to be a voluntary, consensus-driven activity, carried out by and for the interested parties themselves, based on openness and transparency, within independent and recognised SSO’s, leading to the adoption of standards, compliance with which is voluntary. SDOs are often well 36
regulated and transparent organisations which are open to participation. Quasi-formal SSOs are 37
organisations that have attained an important status within a certain area but have not (yet) acquired formal recognition. Examples are the IEEE, the W3C and the IETF. Consortia are smaller groups of likeminded partners that are often ephemeral and established for a specific purpose. Of the 965 SSOs worldwide, the 38 39
majority is a consortium.
§ 1.2: Standards
Standards
According to Dir. 98/34/EC, a standard is a technical specification approved by a recognised standardisation body for repeated or continuous application, with which compliance is not compulsory and which is adopted by an international, European or national standardisation body (see §1.1) and made available to the public. 40
Within this context, compatibility standards (or interoperability standards) are defined as formal or 41
informal agreements regarding how components of a technical system interact. This type of standards is 42
Falke (2000), p. 15
32
For an overview, see Falke (2000), pp. 84-85
33
Falke (2000), p. 1
34
See , inter alia, art. 3(2) and (3) of the NEN Statutes
35
Council Resolution C141/01, 11th Recital
36
Biddle (2012), p. 185
37
Consortia are therefore also referred to as SIGs (Special Interest Groups)
38
This number is derived from the authoritative and frequently updated list maintained by A.
39
Updegrove, to be found at http://www.consortiuminfo.org/links/#.Vj44nsvF_zI . Lastly consulted on 08-12-2015
Art. 1(4) Dir. 98/34/EC and art. 2 (1) Reg. 1025/2012
40
In this article, ‘standards’ refers to compatibility standards
41
Leiponen (2008), p. 1904
specific in nature and requires highly detailed specifications in documents that can contain over 10.000 pages describing the integration of a cumulation of technologies. The main categorical distinction (running 43
parallel to the formal/informal classification) that can be made when talking about standards is de facto and
de jure. The latter category are (i.a.) the ones contained within the definition of Dir. 98/34/EC. De facto
standards are standards that evolve naturally through acceptation by the market (e.g.: VHS, the IBM PC or 44
the Windows OS) while de jure standards are backed by the force of law because they are implemented within some regulatory framework. This framework can either consist of 1) contractual rules provided by an SSO or 2) governmental rules. The majority of standard-setting activities take place within SSOs. Article 45 46
1(9) of Dir. 98/34/EC also makes a distinction between de facto and de jure technical regulations which are compulsory because of direct or indirect inclusion in a regulatory framework. It should be clear that this distinction is not the same as the distinction made above between types of standards. ‘Technical regulations’ are ‘technical specifications’ required by law and by their very nature compulsory. Dir. 98/34/EC thus 47
defines a standard as a subspecies of a ‘technical specification’. This article deals exclusively with formal 48
and informal standards that are governed by a contractual relationship between an SSO-member and the SSO itself. Standards are usually created through coordinated efforts by SSOs and markets are generally
unreliable for developing standards. However, it is possible that a standard is born out of consumer
acceptance. This consumer acceptance can take the form of a very large consumer base in the literal sense 49
or that of a large customer, such as Wal-Mart, which played a major role in defining the Radio Frequency Identification standard. De facto standards can also be imposed by companies with sufficient market-power. For example, IBM once set the standards for the interfaces in the market for plug-compatible mainframes. 50
In addition to the statutory/non-statutory and de jure/de facto classifications made above, ISO maintains a standards catalogue to which the ESO’s and NSO’s adhere. This catalogue contains a hierarchical 51
classification consisting of three levels. The top-level is divided into 40 technical fields of activity in
standardisation. These are then divided into subgroups. The ICS includes the following documents under the term ‘standard’: technical reports, standardised profiles, technical specifications, technical regulations,
Bekkers et al. (2011), p. 2
43
This does not necessarily mean that this is caused by ‘genuine’ acceptation because of the
44
superior quality of a product. Especially in the case of software and hardware, lock-in can be an important driver of market acceptation. For a clear account of lock-in with regard to ICT markets, see Shapiro (1999), chapter 5 and 6 (pp.103 - 171).
Lee (2015), p. 56.
45
Shapiro (1999), p. 237.
46
as defined in art. 1(2) of Dir. 98/34/EC
47
Falke (2000), p. 91
48
Dewatripont & Legros (2013), p. 913-914
49
Peitz & Waldfogel (2012), p. 44-45
50
Falke (2000), p. 95
guides, codes of practice, technology trends assessments and the drafts of these documents. The ESO’s also 52
have their own typology of publications. Depending on the needs of the industry it serves, ETSI publishes 53
the following standards: European standards, ETSI standards, guides, technical specifications, technical reports, special reports and group specifications. European standards may acquire the special status of a 54
harmonised standard where it is published in response to an EC Mandate and aims to meet essential requirements of an EC Directive. An example of a harmonised standard is GSM (EN 301 502) which was published to meet the requirements of the Radio and Telecommunication Terminal Equipment Directive. 55
Types of ETSI standards differ as well in their approval processes and in the obligations they carry. European standards must be transposed into national standards by the associated NSO’s. The approval process is summarised as: 1) approval by a technical committee (comprising of contributing companies and possibly SEP-holders). 2) A public enquiry carried out by the NSO’s. 3) Possible revision based on comments receives after the enquiry. 4) Possible second public enquiry in case of substantial modifications. 5) A final vote (if no comments are received after the second step, this vote is carried out directly) by the NSO’s. Voting is considered to be successful if 71% is in favour. The procedures for the other ETSI publications 56
are similar but shortened versions of this procedure. CEN and CENELEC maintain a similar classification system. Especially with regard to European standards and technical specifications because these are normalised in Reg. 1025/2012. 57
§ 1.3: SEPs
A SEP is a patent that has been declared to an SSO as being essential to a standard, meaning that the implementation of the standard, necessarily infringes the SEP. The main difference between an SEP and a 58
normal patent therefore, is that normally, it is technically possible for an implementer to design around a patent without sacrificing key-functionality. Several studies show that (parallel to the development of more 59
detailed IPR policies) SEPs have become of increased strategic importance in the past few decades. A study
par. 2, ICS
52
Which fall within the ambit of the ICS.
53 http://www.etsi.org/standards/different-types-of-etsi-standards 54 Dir. 1999/5/EC 55 http://www.etsi.org/standards/how-does-etsi-make-standards/approval-processes 56
Art. 2(1)(b) and art. 2(4) respectively.
57
Kuijpers & Kohlen (2012), p. 343
58
par. 52, Motorola
on UMTS (the successor of GSM ), showed that it involved an eightfold increase in patents and a 60 61
threefold increase in patent owners paired with a higher concentration of patents in comparison with its predecessor. The ETSI database now includes 7909 standards and 173621 patents. The majority of SEPs 62 63
are declared in the telecom industry via public networks where the EC identified 4284 SEPs in 2014 of which 244 were litigated (not all within that year). There is no other sector where this much SEP-related litigation occurs. 64
Essentiality
The basic concept distinguishing SEPs from ‘normal’ patents is essentiality. The term ‘essentiality’ is devoid from statutory legal meaning but is nonetheless used in literature, IPR policies and court proceedings to determine if and how a patent becomes essential to a standard. With that, it does play a role in the contractual relation between companies and SSOs. Especially because courts often refer to the definitions given by SSO IPR policies. Kesan & Hayes distinguish three types of essentiality: 1) core essential, meaning that the 65
patent is technologically essential to a core element of the standard, 2) noncore essential, meaning that the patent is technologically essential to an optional function of the standard and 3) commercially essential. 66
Commercial essentiality relates to the situation in which alternative technologies that comply with the standard are unattainable for practical reasons so that they do not represent commercially realistic alternatives. Therefore, this concept is of specific importance to the scope of essentiality. Commercial 67
essentiality might be asserted specifically when it is substantiated by compatibility requirements. The ETSI 68
IPR policy determines that essential means that “it is not possible on technical (but not commercial grounds),
taking into account normal technical practice and the state of the art generally available at the time of standardisation, to make, sell, lease, otherwise dispose of, repair, use or operate equipment or methods which comply with a standard without infringing that IPR.” The CEN-CENELEC Guide defines it as “a 69
patent considered (…) by the patent holder as ‘essential’ when, in his own judgement, it is not possible on technical grounds, considering the state of the art at the time the standard-making process takes place, to
The technology under dispute in the so called ‘smart-phone wars’ between Apple and Samsung
60
Filed as an ETSI harmonized standard under EN 301 502
61
Bekkers et al. (2011), p. 4
62
ETSI IPR database, available at: https://ipr.etsi.org
63
EC Policy brief, p. 3
64
e.g.: paras. 3 - 6, Nokia v. Interdigital 2006 and par. 3, Nokia v. Interdigital 2007
65
Kesan & Hayes (2014), p. 241
66
Updegrove & Bekkers (2012), p. 39-40, an interesting border-case for commercial essentiality
67
would be the slide-to-unlock function on a smartphone Kesan & Hayes (2014), p. 242
68
ETSI IPR policy, art.15(6)
make, sell, lease and otherwise dispose an equipment, product or method which comply with a deliverable without infringing his IPR on that patent”. Apart from the fact that definitions of essentiality among SSOs 70
vary, it is hard to define whether a patent is essential because standards can include a large amount of different technological components. Other questions that arise in the context of essentiality are whether it includes commercial essentiality, other non-essential claims in the same patent, pending patent applications and whether it applies if all other alternatives are patented (because there would be multiple essential technologies). In the case of ETSI, essentiality includes all these components except for commercial essentiality. The scope of essentiality is further defined by the inclusion/exclusion of normatively 71
referenced standards in the standard that is covered by the SEP. Here, the question rises whether the 72
essential IPR covering the referenced standard is also essential to the referring standard. The ETSI policy 73
does not specify this. The definition and scope of essentiality hinges on much more factors than the ones treated here. However, for the purposes of this article it suffices to understand that essentiality is a versatile concept, the definition of which can be relevant to SEP-related litigation.
§ 1.4: Institutions and SEPs: SSO IPR Policies
IPRs can create divergent interests among participants in an SSO. For this reason, SSOs create formal and informal mechanisms in order to streamline the existence of IPRs and the standard-setting process 74
(including the implementation of standards). These mechanisms manifest themselves mainly in IPR Policies. SSO IPR Policies may be considered a form of ‘private ordering’ enabling market participants to collectively contract around the initial entitlements of IPRs. They are multi-party contracts that lay down the rights and obligations of members of an SSO. The value of an IPR policy is thus measured by who is party to the 75
contract, the precision of its wording, the scope of the obligations it imposes and its enforceability. The 76
main overarching goals of any SSO IPR Policy (although these are rarely explicitly mentioned) may be summarised as: transparency in standards development and with regard to SEPs, availability of licenses, preventing patent-hold-up, ambushes or blocking and royalty stacking and preventing discrimination between implementers. An important limitation to the effectiveness of any IPR policy is that it can only be 77
Art. 2.6, CEN-CENELEC Guide 8
70
Updegrove & Bekkers (2012), p. 34-35
71
I.e.: When a standard is built on other standardised technology that is also covered by essential
72 claims. Updergove Bekkers (2012), p. 41 73 in particular patents. 74 Lemley (2002), p. 1909. 75
Updegrove & Bekkers (2012), p. 4
76
P&S Report, p. 38. The list of goals identified in the various research papers is much more
77
binding upon its members. It needs to be noted however, that a standard rarely includes technology 78
patented by a non-member or entity that was not party in any way to the development of a standard. The 79
two core elements included in SSO IPR policies are 1) search and disclosure and 2) licensing rules. 80
Search and disclosure
Regulating disclosure is a balancing exercise. A policy that restricts disclosure too much will create uncertainty regarding the existence of essential IP, whereas very broad disclosure provisions may lead to suspicion of concerted practices. The ETSI IPR policy requires members to use ‘reasonable endeavours’ to 81
inform the SSO of essential patents ‘in particular during the development of a standard’ and in a ‘timely
fashion’ and on a ‘bona fide’ basis. The clause pursuant to that explicitly rules out any obligations to 82
conduct searches. Disclosures of essential IPR are made in an information statement and can be combined 83 84
with a licensing declaration (treated below). Similarly, the CEN-CENELEC IPR policy requires that ‘any
party participating in the work of CEN and CENELEC’ draws attention to any patent or pending patent
application ‘as early as possible during the development of the deliverable’ to the best of their knowledge and in good faith. The policy also rules out a requirement for patent searches. This is common practice 85
outside of Europe as well. The disclosure of patents by members of an SSO is an important aspect of the 86
standardisation process. First, in its informational and non-binding form. In this form it is also known as a 87
‘patent call’ which is conducted at the start of each meeting of a working group that is in the course of developing a standard. This type of disclosure is important so that the working group knows whether it can design the around the patent. Moreover, if a member acts in bad faith, patent calls make it harder to later
Updegrove & Bekkers (2012), p. 32
78
Updegrove & Bekkers (2012), p. 33. Companies can also be contractually bound to an SSO
79
through individual contracts instead of membership. That is why the phrasing ‘in any way’ is included here. This article will not cover the colourful variety in which companies may be bound to an SSO.
Rysman & Simcoe (2008), p. 1923. Essentiality is also a central element but is treated in par. 1.3
80
Staniszewski (2007), p. 669
81
Art. 4.1, ETSI IPR policy
82
Art. 4.2 ETSI IPR policy
83
Appendix A to the ETSI IPR policy, p. 42
84
Art. 3, CEN-CENELEC Guide 8
85
See for example, art. 6.7 in the W3C patent policy.
86
That is to say: it does not create a licensing commitment, the duty to disclose during a patent call
87
is binding. This is specified in the ETSI and CEN-CENELEC policies in the words: in particular
during the development of a standard’ and ‘as early as possible during the development of the deliverable’ respectively.
claim that it was unaware of the duty to disclose. The second form of disclosure is the formal one which is 88
done in the information statement or in conjunction with the licensing commitments. The scope of formal disclosure is never absolute. It is defined by the wording of the SSO IPR policy which aims at 1) the assumed knowledge of the organisation (or representative) and 2) IPR that is essential, potentially essential or may be(come) essential. The ETSI policy specifies the assumed knowledge with the words reasonable 89
endeavours’’ and the IPR to be disclosed with the words ‘may be or may become’. The CEN-CENELEC 90
IPR policy specifies the assumed knowledge with the phrase ‘to the best of his knowledge’ and the IPR to be disclosed with the words ‘any known patent or any known pending application (…) that (…) may be
considered as an essential patent for the deliverable’. 91
SSOs do not evaluate the essentiality of a patent, nor are members of SSOs generally required to conduct patent searches in order to reveal potentially essential patents. Moreover, disclosure of essential patents by 92
a member to an SSO occurs mainly on the basis of self-assessment. There are no checks on the quality of 93
that assessment. This absence brings with it the risk of over-claiming by opportunistic patent holders but is 94
nonetheless standard practice within and outside of Europe. 95
Licensing and FRAND
Licensing SEPs occurs through a great variety of mechanisms ranging from patent pools , cross-licenses, 96
technology auctions to non-assertion commitments. In the following, these specific mechanisms are left 97
aside and the SSO-model is discussed. The approaches takes by SSOs can be categorised into (F)RAND 98
RF , (F)RAND and RAND-Z. These approaches are different flavours of licensing commitments made by 99 100
Updegrove & Bekkers (2012), p. 49
88
Updegrove & Bekkers (2012), p. 54, other phrasings are possible.
89
Respectively: art. 4.1 and Appendix A under ‘IPR Information statement’ of the ETSI IPR policy
90
Art. 3, CEN-CENELEN Guide 8
91 Lee (2015), p. 62 92 P&S Report, p. 114 93 P&S Report, p. 146 94
See for example IETF RFC 3979: ‘(…) the IETF can make no actual determination of validity,
95
enforceability or applicability of any particular IPR claim (…).
Such as the MPEG LA patent pool
96
Blind et al. (2011), p. 114
97
Acronym adopted in the legal practice referring to the often used terms in an IPR policy, stands
98
for fair, reasonable and non-discriminatory. Royalty-free
99
zero-royalty. Categories are derived from Blind et al. (2011), p. 112
the SEP-holder, which SSOs use to secure guarantees for the availability of licenses for standardised
technology. Constraining licensing terms by imposing special obligations also aims at limiting the substantial market power necessarily enjoyed by SEP-holders. RF is a commitment that is completely royalty-free. 101
RAND-Z is a commitment where the licensor can not make any money off the license but retains its right to stop or impede production of the licensed products. The approach of the SSO is dependent on the nature of 102
its technological area and its members. For example, members might be SEP-owners whose business 103
models are based on the sales of products that implement standardised technology. An RF or FRAND-Z scenario would be in their best interests. Also, technological areas with strong networking effects are more 104
reliant on interfaces to create interoperability, making it less probable that standards are proprietary. Exemplary are the W3C, whose only licensing option is RF, and the IETF that explicitly strives to create 105
standards with ‘no known IPR claims or, for technologies with claims against them, an offer of royalty-free
licensing.’ Both are SSOs that create internet-standards, an area most susceptible to networking effects. 106
Smaller SSOs usually bind their members to licensing commitments through a participation based model, in which FRAND is a part of the IPR policy which is binding through membership. The European SSOs use the commitment-based model, in which separate declarations are made that impose FRAND-obligations on the declarant. In ETSI, commitments are made through a declaration form which is attached to the IPR 107
policy. The SEP holder provides a binding commitment to license any essential IPR on fair, reasonable and 108
non-discriminatory terms. CEN and CENELEC seek these commitments in a similar way. However, 109 110
they provide the option to license on FRAND terms or free of charge with further negotiations left to the parties outside of CEN or CENELEC. In both cases, the licenses are irrevocable and implementers are subject to reciprocal obligations, meaning that any essential IPR in their possession must be licensed to the declarant as a cross-licensing scenario. Also, both policies leave the option to refuse to license in the declaration form. Neither ETSI nor the CEN-CENELEC policy specifies the obligations arising out of FRAND. Nor do they provide guidelines regarding the duration or royalty-fee for the license. The
vagueness surrounding FRAND can be interpreted as a limitation to the contractual enforceability of SSO IP
Maskus & Merrill (2013), p. 54
101
http://www.goland.org/buyingopenstandards/
102
After all, the IPR policy is also drafted and approved by the SSOs members.
103
Maskus & Merrill (2013), p. 52
104
Art. 5(5), W3C patent policy
105
Art. 8, first paragraph, RFC 3979
106
The difference between participation-based and commitment-based IPR policies is taken from
107
P&S Report, p. 40
Appendix A to the ETSI IPR Policy
108
Art. 6.1 of the ETSI IPR Policy.
109
See declaration form attached as Annex 2 to the CEN-CENELEC Guide 8
Rules. Practice has proven, however, that courts are not hesitant to determine what constitutes a royalty on 111
FRAND-terms. Moreover, FRAND has become a tool for courts to evaluate and regulate the behaviour of 112
companies that are party to SEP licensing negotiations. 113
Lemley (2002), p. 1914
111
Discussed further in par. 2.2
112
Famous European cases are the Orange Book standard-case of the BGH and more recently
113
Chapter 2: Delineating Enforcement
§ 2.1: Enforcement of SEPs in the EU
IMAGE 1.1 114
Image 1.1 provides an overview of the interaction between the concepts elaborated in chapter 1 and the legal mechanisms described in the current chapter. Three preliminary conclusions may be drawn from this image. First, there is no uniform framework governing SEPs or any legal instrument that directly addresses them. 115
Second, the main legal instruments governing SEPs are contract law (in the shape of SSO rules) and antitrust law. Third, the enforcement of SEPs is effectuated through ‘regular’ enforcement regulations. The latter 116
Edited image taken from P&S-report, p. 31. The paragraphs and chapter refer to where the
114
components are treated in this article. The boxes marked with an asterisk are not treated in this article because they are not of specific importance to the enforcement of SEPs. Although patent pools have frequently been proposed as a solution to issues in standardisation, they are not part of the discussion here. Patent offices may contribute to the quality of SEPs through collaboration with SSOs but are not directly involved in the enforcement of SEPs. The EU framework for
standardisation is an important legal tool in regulating SSOs and by extension SEPs. However, it is not directly relevant for the enforcement of SEPs, as can be inferred from the image.
Lemley calls it a ‘hodgepodge of legal theories, each with their own shortcomings’, Lemley
115
(2002), p. 1936. Kahin (2007) explains that ‘there is no recognised domain for standards law or
standards policy’.
Although this is not explicitly mentioned in the image, it may be inferred because SEPs are,
116
from a purely legal viewpoint, regarded as ‘regular’ patents. I.e.: they do not have a special status in statutory law.
are briefly set out in this paragraph. Antitrust scrutiny with regard to essential facilities is treated in §2.2. SSO rules have been introduced in §1.4 and their role within antitrust is discussed in §2.3. The fact that antitrust rules and SSO IPR rules are intertwined can be seen from the fact that the EC sets requirements for SSO IPR policies in order for them to be better compatible with antitrust laws. In other words: provisions 117
in SSO IPR policies are used by antitrust authorities and courts to substantiate and interpret antitrust requirements when applying them to SEPs.
Actions
In principle, the EC makes no difference between patent holders and SEP-holders, meaning that both can 118
have a legitimate recourse to the array of actions described by Dir. 2004/48/EC. It incorporates civil law 119
measures under the TRIPS Agreement into the EU legal framework and intends to harmonise the measures 120
available to IPR holders. The actions are divided into actions relating to 1) acquiring and preserving 121
evidence, 2) the right to information of the claimant in the context of proceedings, 3) provisional and precautionary measures,4) corrective measures (such as injunctions and pecuniary compensation) and 5) damages. These actions apply to all infringements of IPRs protected under European and national 122
legislation
General r
estrictionsArticle 3(1) of Dir. 2004/48/EC incorporates a notion of fairness and equity in assessing the use of the actions provided. Furthermore, the Dir. explicitly states that the application of its rules should ‘not affect the
application of the rules of competition, and in particular articles 81 and 82 of the Treaty.’ This inherent 123
outline of the boundaries of Dir. 2004/48/EC serves as a base for restricting its use in cases where it would amount to an abuse. Similarly, the TRIPS Agreement provides for the possibility of adopting ‘measures
necessary to (…) promote the public interest in sectors of vital importance to their socio-economic and technological development (…)’ and ‘appropriate measures (…) needed to prevent the abuse of intellectual property rights by right holders or the resort to practices which unreasonably restrain trade or adversely affect the international transfer of technology.” TRIPS further provides that Members may specify ‘in 124
These requirements include a balanced and adequate IPR policy which includes good faith
117
disclosure and FRAND clauses in order to ‘increase the likelihood that the implementers will be
granted effective access to the standards elaborated by that SSO’, paras. 284-286, EC Guidelines.
MEMO/13/403, p. 1 118 par. 278, Motorola 119 COM (2010) 779, p. 5 120
Recital 8, Dir. 2004/48/EC
121
Articles 6 to 13, Dir. 2004/48/EC
122
Recital 12, Dir. 2004/48/EC
123
Respectively, art. 8(1) and 8(2), TRIPS
their legislation practices and conditions that may in particular cases constitute an abuse of intellectual property rights having an adverse effect on competition in the relevant market’. The following elaboration 125
represents how European authorities and courts have used these provisions to restrict the use of Dir. 2004/48/ EC.
§ 2.2: Antitrust scrutiny
126The EC maintains that ‘standardisation may lead to anti-competitive results by preventing certain companies
from obtaining effective access to the results of the standard-setting process (that is to say (…) the essential IPR for implementing the standard.’ Consequently, evaluating the use of SEPs under antitrust law occurs 127
mainly in the context of the refusal of access to an essential element of the productive process. The refusal to license patents relating to interoperability may constitute an abuse of a dominant position, as defined in 102 TFEU, where such information is indispensable for the exercise of an activity in neighbouring or related markets. The case law relating to article 102 TFEU represents a recognition of the CJEU of a ‘special 128
responsibility’ of a dominant firm towards the competitive process in the market which it ‘dominates’, 129
requiring the firm to not impair genuine undistorted competition on the common market. The obligation is a norm for the commercial conduct of the firm and is restrictive depending on the amount of competitors in the relevant market and the size of the firm’s market share. A specific manifestation of the special responsibility is the duty to deal with or supply customers that are also competitors. Restraints are imposed on the firm’s freedom to refuse to deal with competing firms. As the wording of article 102 TFEU suggests, dominance 130
is a question that is prior to an evaluation of the refusal to deal.
The ECJ defines dominance as ‘the possession of economic strength which enables a business to prevent effective competition in the relevant market by affording it the power to behave, to an appreciable extent, independently of competitors, customers and consumers. 131
The “essential facilities” doctrine then, is a more specialised sub-species of the dominant firm’s duty to deal with its existing competitors (as described in §2.3). The ‘classic’ case for invoking an essential facilities argument is when a competitor seeks physical resources controlled by the dominant firm that are needed to
Art. 40, TRIPS
125
A side-note here is that this article will not treat liability of SSOs when they act as a cartel or in
126
relation to horizontal agreements. See for example recital 13 of Reg. 1025/2012 This article deals exclusively with antitrust-liability of SEP-holders.
par. 268 EC Guidelines
127
Pereira (2011), p. 177
128
The scope of which is determined by the specific circumstances of the case and at the time
129
when the abusive conduct takes place, see par. 272, Motorola Ong (2005), p. 216
130
Staniszewski (2007), p. 676
compete with it. The duty to grant access to an essential facility is strongest where it can not be feasibly replicated by the competitor on practical, economic. environmental or other grounds. 132
Physical essential facilities
An early elaboration of the essential facilities doctrine can be found in Volvo v. Veng, in which the refusal to license design rights in car spare parts created a barricade to trade in a downstream market. The CJEU 133
affirmed that the exercise of an IPR can be an abuse of a dominant position where the exercise can not be justified by any rational purpose, thereby artificially shielding a product from competition. However, to 134
convert a ‘mere’ refusal to license into an abuse of a dominant position, it would have to be accompanied by certain abusive conduct. The leading judgment that further implemented the doctrine in the EU is the 135
Bronner-case, which effectively relied on the Magill-case . In Bronner, the plaintiff sought access to a 136
home-delivery system by alleging that Mediaprint infringed 102 TFEU because of their refusal of access to that system. The CJEU first considered that Mediaprint held a de facto monopoly because it operated the only home-delivery scheme in Austria. If it were established before the national court that 1) a separate market for home-delivery schemes existed and 2) that there was an insufficient degree of interchangeability between Mediaprint schemes and other schemes, Mediaprint would be a dominant firm under 102 TFEU. 137
The considerations that led the CJEU to conclude that the scheme was not an essential facility, even if it was the only scheme present and held by a dominant firm, were 1) less advantageous methods of distribution existed and were used by other publishers, 2) there were no technical, legal or economic obstacles for using those alternatives. The evaluation of a refusal of access to physical essential facilities under 102 TFEU was also carried out in Commercial Solvents and Telemarketing. In these cases, the refusal to supply raw materials to downstream manufacturers and the refusal to allow independent telemarketing undertakings television space was judged to constitute ‘abusive conduct’ under the essential facilities doctrine. These 138
cases could be considered the ‘physical’ precedents for extending the application of the essential facilities doctrine to facilities that are protected by IPRs.
Ong (2005), p. 217
132
I(a), Volvo v. Veng
133
p. 6219, Volvo v. Veng: The rational purpose being the reward of those who, by creative effort
134
and/or work and/or investment have brought on to the market products that are new and useful or attractive.
Ong (2005), p. 220
135
The Bronner-case equivalent for essential facilities protected by IPRs, the two cases are
136
closely related. par. 35, Bronner
137
Respectively: par. 25, Commercial Solvents and par. 26, Telemarketing
IPR-encumbered essential facilities
In the same way, refusal of access to essential facilities that are protected by IPRs can qualify as an abuse under 102 TFEU. The Magill-case affirmed that the refusal to grant a license (which is the IPR-equivalent of granting access to physical facilities) by a dominant firm in itself can not constitute an abuse under 102 TFEU but that the exercise of an exclusive right by a proprietor may, in exceptional circumstances , 139
constitute an abuse of a dominant position. The exceptional circumstances were found in 1) that the 140
product was indispensable to carrying on the business in question , 2) the refusal prevented the appearance 141
of a new product for which there was potential consumer demand, 3) there were no objective considerations for justifying the refusal and 4) the refusal was likely to exclude all competition in the relevant market. 142
Although closely related through the normative framework that was applied, Bronner and Magill are fundamentally different in that the latter treats access to licenses. This means that a dominant firm can be dominant in a market in a ‘physical’ sense (mostly on upstream markets) while retaining IPRs that are essential in (e.g.) a downstream market. Also, not the physical assets of the company are essential but the IPRs themselves must be indispensable and fulfil the first element mentioned above. However, a firm 143
‘physically’ operating on an upstream market (or operating on both upstream and downstream markets) while retaining IPRs that may distort competition on a downstream market is not excluded from the possibility of qualifying under 102 TFEU. This is also referred to as a ‘vertically integrated company 144
abusing a dominant position on an upstream market to restrict competition on a downstream market’. Continuing further down the track of evaluating the exercise of IPRs under 102 TFEU, IMS v. NDC involved the refusal to grant access to a copyrighted brick structure used for recording and presenting sales data and that had turned into a de facto industry standard because customers would not accept data in a different 145
format. The decision was preceded by an order to license issued by the EC based on the existence of the ‘exceptional circumstances’ in the tradition of Volvo v. Veng and Magill. The German court sought to know whether the following elements were important to establishing exceptional circumstances that could convert
Also named the ‘exceptional circumstances doctrine’ by Staniszewski (2007), p. 666
139
paras. 49 - 50, Magill
140
i.e.: that there is no actual or potential substitute in existence for the facility or that it is
141
impossible to implement such an alternative because of technical, legal or even economic means. It is not enough to argue that it is not economically viable to use an alternative. (par. 41-46 in
Bronner)
par. 40, Bronner
142
Ong (2005), p. 218
143
“The mere fact that input essential for the production of downstream goods or service is not
144
independently marketed by the dominant undertaking does not operate to exclude the unjustified refusal of access to that input from constituting abusive conduct. [because it] imposes a significant restriction of competition in the market for downstream goods.” par. 44, Opinion of the AG in IMS v. NDC
Caused, inter alia, by the free distribution of the structure to pharmacies and doctors’ surgeries
145
a ‘refusal of access’ into ‘abusive conduct’: 1) the fact that the consumer base does not accept another format. 2) The extent to which people from other parts of the market were used to develop the copyrighted product. 3) The switching costs involved if consumers want to start using the competing product. The 146 147
CJEU (referring to the AG) ruled that consumer acceptance was premised on the indispensability of the product in having access to a market and should therefore be answered by fulfilling the Magill-criteria. 148
The latter two elements were to be taken into consideration in evaluating whether a refusal to license constitutes abusive conduct. In particular because the participation of the users in the development of the 149
product and the switching costs could be indications for the existence of dependency (particularly at a 150
technical level) of those users on the product. 151
§ 2.3: The role of SSO IPR policies
In this paragraph, the three basic elements of SSO IPR policies as set out in §1.3 and §1.4 will be discussed. Essentiality is an element that is strongly related to the scope and validity of a patent and therefore of a more technical nature. When invoked as a defence, non-essentiality has a determinative capacity because it is a threshold criterium for the applicability of FRAND and other SSO IPR policy provisions. It is therefore treated first. Breaching disclosure and FRAND obligations may give rise to several types of (contractual) liability. However, these will only be discussed in the context of antitrust proceedings. 152
Essentiality
As noted in par. 1.4, there is no control on the assessment of essentiality outside of the courtroom, leading to a practice in which the validity and essentiality of an SEP is not certain. A court can decide on 153
the original wording of the court is ‘material outlay’, meaning that the definition of costs is more
146
ample than just the financial dimension. par. 17, IMS v. NDC 147 par. 22, IMS v. NDC 148 paras. 28 - 30, IMS v. NDC 149
For example, a product that is completely customised to the needs of the user and that is
150
developed in dialogue with that consumer. One could think of specific software that is developed for a company with a specific functionality adapted to the internal workings of the company.
see more specifically par. 29, IMS v. NDC
151
Contractual liability has no direct relevance for the points made in Chapter 3, which form the
152
apex of this article. P&S Report, p. 147
infringement and therefore the essentiality of a product, with the former necessarily preceding latter. 154 155
Until such a decision has been made, an SEP (like a ‘normal’ patent) is presumed to be essential. 156
Essentiality is a technical question that arises in addition to the usual legal validity of a patent. Affirmative 157
answers to these questions are not mutually exclusive. A patent can be valid but nonetheless declared non-essential to a standard by a court. Conversely, a patent can be invalid but some claims might still be 158
declared essential to a standard by a judge. In these cases, a court will deal with the question of 159
essentiality on the assumption that the patent is valid. Strictly speaking however, if a patent is invalid, it can not be an SEP because it can not be essential to any standard. Depending on the reason for it, non-validity 160
might be cured by an amendment. If the amendment is accepted and does not make the patent non-essential, the patent will be an SEP and can be asserted against implementers. It is possible that an implementer challenges infringement (and therefore essentiality) as a defence against an injunction. This is also called a declaration of non-essentiality. In IPCom v. HTC, the defendant sought a declaration of non-infringement 161
as a defence against an injunction brought by IPCom. The latter claimed that HTC, an implementer of the UMTS standard, had infringed patents that IPCom had declared essential. The trial was of a technical nature and non-infringement was found, which implied the non-essentiality of IPCom’s SEPs. In Nokia v. 162
Interdigital, the plaintiff, after having acquired a license for patents declared essential to ETSI for the 3G
standard by Interdigital, alleged that the licensed patents were non-essential. Nokia reserved the right to
It is assumed here, similar to the approach of the EC in Motorola and the UK High Court in
154
IPCom v. HTC (see paras. 12(iii) and 121), that the question of essentiality will be a corollary to the
question of infringement. I.e.: if it turns out in a court proceeding that a patent’s scope covers a part of the standardised technology that is essential to it and is therefore infringed by the
implementation of the standardised technology, essentiality is proven. This approach is in line with the ETSI definition of ‘essential’ IPR because it makes it dependent on the infringement of the IPR. See art. 15(6) of the ETSI IPR policy.
par. 28, Nokia v. Interdigital: “If apparatus in accordance with the standard does not infringe,
155
that will be the end of it unless there is a relevant “implementation-specific” feature of a compliant system. But nothing can be notified as technically essential if it relates merely to implementation-specific features.”
par. 61, Motorola
156
Dewatripont & Legros (2013), p. 915
157
E.g.: in IPcom v. HTC, non-infringement was found, resulting in non-essentiality. The patent,
158
however, was valid.
e.g.: paras. 287 - 292, Nokia v. IPCom 2010.
159
par. 167, Nokia v. IPCom 2012
160
Blind et al. (2011), p. 150
161
paras. 2 and 121, IPCom v. HTC
challenge the validity of the patents found to be essential. Following a technical assessment in which the 163
claims of the patents were held against the 3G-standards, three of the four patents were declared non-essential by the court. This case shows that an implementer can also seek a declaration of non-non-essentiality 164
without a potentially infringing product. If essentiality is proven, a validity defence might be a (last) 165
option. Stevenson & Jelf derived the following approach from Nokia v. Interdigital: 1) Hold the patent against the relevant standards, 2) identify the correct construction of the patent, independently of the
standards and through the eye of the skilled person, 3) consider to what extent the correct claim construction corresponds with the standards, keeping in mind where the claims retain their own meaning and where they have to be stretched to fit the standard, 4) decide whether what is required by the standards falls within the language of the claim as understood by the skilled person. The assessment of essentiality in Nokia v. 166
IPCom 2012, where Nokia challenged the essentiality of a patent owned by IPCom to the UMTS , GSM 167 168
and DECT standards, was constructed in the same way. It should be clear that the outcome of an 169 170
affirmative declaration of essentiality in court is never a ‘yes-or-no-question’. Some SEPs are of a higher quality than others (because of forward citation counts or technological value) or may be situated closer to the core functioning of the standard at issue. Also, essentiality functions as a gatekeeper for the SEP-status 171
of a patent and should therefore not be automatically presumed when a holder asserts his patents. SEP-based actions should be preceded by evidencing essentiality, as was made clear in a recent judgment by the Court of first instance of Paris. 172
par. 6, Nokia v. Interdigital 2007, the mention of this reservation serves to illustrate the relation
163
between validity and essentiality. Challenging validity will only be useful for a defendant if the patent is found to be essential to the standard concerned. If the patent is non-essential, it is not infringed and a validity-defence will unnecessarily prolong the proceedings.
par. 134 Nokia v. Interdigital 2007
164
Stevenson & Jelf (2008), p. 458
165
Stevenson & Jelf (2008), p. 459
166
Universal Mobile Telecommunications System
167
Global System for Mobile communications
168
Digital Enhanced Cordless Telecommunications
169
For the assessment of essentiality as identified by Stevenson & Jelf under the 4th element as
170
listed above, see paras. 166 - 186, Nokia v. IPCom 2012 Kang & Motohashi (2014), p. 34
171
http://www.limegreenipnews.com/2015/05/france-sep-cases-begin-by-evidencing-essentiality/
Disclosure
If a patent is essential, the rules of an SSO IPR policy regarding licensing and disclosure apply. Breach of disclosure obligations can contribute to liability in the form of 1) an abuse of a dominant position or 2) 173
constitute fraud. Additionally, intentional delay, caused by an ETSI member that deliberately withholds 174
IPR disclosures, can be sanctioned by the General Assembly of ETSI. Only disclosure in the context of the 175
abuse of a dominant position will be discussed here. Concepts like ‘patent-holdup’ and ‘patent ambush’ are related to anti-competitive disclosure-behaviour of SSO-members. Hold-up relates to a situation in which company A is in some way dependent on the input from company B. After having invested in the input from company A, company B starts to assert or levy fees for patents that have not been disclosed in earlier 176
negotiations. Hold-up in the case of SEPs would be a variety where the dependency is caused by the existence of an industry standard, compliance with which is crucial to access to a market. Disclosure 177
during the adoption of a standard is especially relevant because it is very probable that a non-dominant company becomes dominant after the adoption of a standard. If dominance can reasonably be foreseen at the time of the adoption of a standard by a company that breaches its disclosure requirements, the non-disclosure could be seen as an abuse of an (anticipated) dominant position. In Rambus, an early landmark-case 178
concerning the disclosure behaviour and royalties levied for DRAM-chips, standardised by JEDEC with Rambus being a member of the working group, the EC considered that Rambus deliberately refrained 179
from disclosing its patents while participating in the standardisation process , thereby breaching the duty of 180
good faith. Especially in view of the fact that there were alternative technologies available which would’ve enabled JEDEC to design around the IPR, had the patents been disclosed. The EC spoke of ‘intentional deceptive conduct’ and considered the breach of the disclosure-obligations as an element that made the 181
levy of royalties incompatible with art. 102 TFEU. Apart from the Rambus-case, non-disclosure does not often play a major role in antitrust proceedings in the EU because it contributes mainly to the acquisition of a dominant position, which is not punishable in Europe. Remedies dealing with this type of breach are
Non-exhaustive
173
Barazza (2014), p. 562
174
Art. 14, ETSI IPR Policy and par. 2, note 2 and 3, ETSI IPR Guide
175
When it asserts undisclosed patents, the term used is ‘ambush’. When it starts levying
176
unexpectedly high royalties, the term is ‘hold-up’. As far as the author’s concerned, they are both species of the same genus, i.e.: abusive deployment of undisclosed IPR or undisclosed licensing terms. Haber et al. (2015), pp. 7-8 177 Staniszewski (2007), p. 676 178 COMP/38.636 - Rambus 179
Par. 40 and 41 in Rambus
180
This conduct was caused in part by the way that Rambus had participated in the SSO, not
181
therefore mainly found in civil law. Lastly, if the existence of a non-disclosed patent comes to the attention of an ESO after it adopts a standard, it can seek the withdrawal of the recognition of the standard under Dir. 94/34/EC. 182
FRAND
When a patent is disclosed to an SSO and FRAND commitments have been made, generally the role of the SSO is finished. The enforcement of FRAND commitments, therefore, is usually instigated by either an 183
implementer who finds himself the subject of a SEP-based injunction or 2) competition authorities and courts that use it to substantiate abusive conduct in antitrust proceedings. Perspectives on FRAND range from interpretations holding that it implies refraining from using injunctions to just negotiating in good faith. In EU case law, the breach of FRAND-commitments can give rise or contribute to several types of 184
liability for the SEP-holder including (non-exhaustively): 1) violating the pre-contractual good faith , 2) 185
breaching contractual obligations towards the SSO and the potential licensee as a third-party beneficiary, 3) abuse of a dominant position, 4) patent misuse. Again, this article will only discuss how FRAND operates 186
in the context of antitrust proceedings. In these proceedings, the concept of FRAND manifests itself in two ways. First, as a commitment to license on FRAND-terms. Second, as a norm for the substance of a licensing agreement (i.e. royalty rates, dispute resolution, (geographical) scope of license, cross-licensing clauses). 187
With regard to the commitment, European caselaw has crystallised FRAND into a series of steps that have to be taken by the SEP-holder (before seeking an injunction) in order to ensure that the implementer was not ‘objectively willing and able to acquire the license’. If this is not the case, an injunction sought by a SEP-188
holder can be abusive. National courts maintain different perspectives regarding the precise conditions for 189
willingness. An investigation by Barazza produced the following common elements: 1) an offer containing the determination of royalty rates (incl. payment/deposit of royalties for present/past use), 2) the withdrawal of pending invalidity actions, 3) active negotiations regarding royalties, 4) an expression of willingness to enter into negotiations and 5) an offer circumscribed by geographical limitations with regard to the SEP portfolio. In Motorola, an injunction was sought and enforced for the GPRS standard while Motorola had 190
Staniszewski (2007), pp. 676 - 677 182 P&S Report, p. 42 183 Barazza (2014), p. 552 184
Which was discussed in the Dutch case Samsung v. Apple.
185
Barazza (2014), p. 561
186
par. 40, Opinion of the AG in Huawei v. ZTE
187
Nihoul (2015), p.1
188
paras. 73 - 74, Opinion of the AG in Huawei v. ZTE
189
of which the essentiality is not a point of dispute between the parties involved. If a part of the
190
committed itself to comply to the requirements set by the ETSI IPR policy. The EC considered that the ‘exceptional circumstances’ required for qualifying under 102 TFEU were embodied in the commitments 191
made by Motorola and the fact that non-compliance with these commitments would endanger benefits of 192
the standard-setting process such as increased compatibility, interoperability, competition, lower production and lower sales costs. The EC explained the commitment as a choice of the SEP-holder to monetise the 193
standard essential technology without excluding any implementers of the standard concerned, creating (in the implementers) the reasonable expectation that this will occur. Samsung, an EC antitrust-decision 194
concerning injunctions against Apple for the UMTS standard, evaluated the enforcement within the same framework as Motorola, including the commitment to license on FRAND terms as an element of the 195
exceptional circumstances. However, it also evaluated the unwillingness of Apple as a potential objective justification for the conduct of Samsung, reaching the conclusion that this was not the case, Samsung was 196
obliged to refrain from seeking injunctions against willing licensees. Following critique on Motorola and 197
Samsung highlighting the vagueness of the concept of willingness , Huawei v. ZTE provided some clarity 198
as to what constitutes FRAND-negotiations. The reference for a preliminary ruling concerned unsuccessful licensing negotiations after which Huawei sought an injunction against ZTE for base-stations made with 199
LTE-technology. ZTE raised an abuse of a dominant position as a defence. The CJEU explicitly mentioned 200
that the case should be judged based on the caselaw set out above in §2.2, but that the circumstances are different than the ones that gave rise to that caselaw, because of the standardisation context. The CJEU 201
interpreted FRAND as a justification for the imposition of specific requirements on the SEP-holder. The 202
SEP-holder should 1) alert the alleged infringer of the infringement, designating the patent and specifying the infringement and 2) after the infringer has expressed its willingness to conclude a licensing agreement on FRAND terms, present to the infringer a specific, written offer for a license on such terms, specifying the
As discussed in §2.2. 191 par. 1, Motorola 192 par. 284, Motorola 193 par. 294, Motorola 194 section 4.4.1, Samsung 195 section 4.4.3, Samsung 196 par. 122, Samsung 197 Nihoul (2015), p. 1 198
An injunction prohibiting the infringement, the rendering of accounts, the recall of products and
199
an award of damages.
paras. 21 - 27, Huawei v. ZTE
200
par. 48, Huawei v. ZTE
201
par. 59, Huawei v. ZTE