A Commentary on WIPO's The Management of Internet Names and Addresses: Intellectual Property Issues

A. Michael Froomkin
Professor of Law & Quondam "Public Interest Representative" to the WIPO Panel of Experts
University of Miami School of Law

Laina Raveendran Greene, Member WIPO Panel of Experts
has authorized me to note her agreement with the substance of this Commentary.

Harald Tveit Alvestrand, Member WIPO Panel of Experts,
has posted his own "comments to the WIPO process and its result"

May 19, 1999 version 1.0a [typographic changes only]

Please check http://www.law.miami.edu/~amf/commentary.htm for updated versions of this document.


Executive Summary

The World Intellectual Property Organization's Final Report on "The Management of Internet Names And Addresses: Intellectual Property Issues" is in many respects a substantial improvement on WIPO's Interim Report, RFC 3.

Summary of Changes From RFC 3

My "Critique of RFC 3", issued as a personal commentary during the comment period on the Interim Report, identified the following major areas of concern regarding the Interim Report:
  1. Bias. The plan was biased in favor of trademark holders;
  2. Excessive scope. Rather than addressing only trademark law, and focusing on the "cybersquatting" problem, RFC 3 sought to create a system that could adjudicate every type of intellectual-property dispute relating to a domain name-including disputes based on controversial theories such as the right of personality.
  3. `Smorgasbord' approach to law. Instead of directing arbitrators to apply otherwise applicable law, RFC 3 proposed using additional, different, rules it selected-rules that will often disadvantage registrants.
  4. Enabling censorship. The RFC 3 failed to protect fundamental free-speech interests including parody, and criticism of corporations;
  5. Potential for intimidation. RFC 3 created an expensive loser-pays arbitration process with uncertain rules that would intimidate persons who have registered into surrendering valid registrations thus enabling increased "reverse domain name hijacking";
  6. Relied on potentially unenforceable contracts. Because the contracts of adhesion proposed by WIPO were so one-sided, there was reason to believe they would not be enforceable in court and that the entire proposal was therefore unworkable.
  7. The nature of available judicial review. RFC 3 would have allowed disappointed challengers to domain names registrations to appeal to a court in all cases, but would often deny this privilege to the original registrant if he lost.
  8. Zero Privacy. RFC 3 provided zero privacy protections for the name, address and phone number of individual registrants;
  9. Discussion of new gTLDs. RFC 3 took an over-timid approach to new gTLDs. The Final Report's conclusions are welcome, as far as they go - but do not go far enough.
  10. Treatment of famous marks. RFC 3 created new, cumbersome, unwarranted, procedures to protect a potentially unlimited number of "famous" trademarks.
Many of the above issues have been wholly or substantially addressed in the Final Report. Although the seventh has not been addressed, the ill effects will be lessened (although not eliminated) by the reduced scope of the revised dispute resolution procedure. The eight and ninth issues have been, effectively, put on ICANN's shoulders. After some fine-tuning, especially regarding the procedural timetable, the anti-cybersquatting provisions will represent a significant improvement over the current NSI dispute policy-which should benefit both trademark owners and legitimate domain name registrants. Unfortunately, the Final Report makes no material changes in the proposed special privileges for a new class of globally famous marks.

Several important issues raised for the first time in the Final Report require careful consideration, notably the definition of cybersquatting, and the proposed procedural timetable. As the Final Report contains a wealth of material that is new, or substantially different from the Interim Report, including the critical Annexes, further review and public comment are likely to be essential before ICANN takes action.


  1. The World Intellectual Property Organization's Final Report on "The Management of Internet Names And Addresses: Intellectual Property Issues" (hereinafter, "Final Report") is in many respects a substantial improvement on the Interim Report. However, because the Final Report contains a wealth of material that is new, or substantially different from the Interim Report, including the crucial Annexes, further review and public comment is likely to be essential before ICANN takes action.
  2. Preface: A Personal Analysis

  3. On September 28, 1998 I was asked to serve on a World Intellectual Property Organization "Panel of Experts" that WIPO had formed to advise it on its forthcoming recommendations relating to domain name/trademark issues. The WIPO official who invited me to join the panel said I had been selected to ensure that there was a "public interest" advocate inside the WIPO process. I accepted on condition that I be allowed to speak freely at all stages of the process, and was assured that this condition presented no difficulty.
  4. I should emphasize that what follows is a strictly personal analysis. My attendance at WIPO regional consultations have led me to understand that not everyone comprehends how WIPO has chosen to employ the Panel of Experts. Our limited role is described in more detail below. In this process, the WIPO staff drafts the documents WIPO issues (and the document is not subject to prior review by the member states); the Panel of Experts merely makes substantive or editorial suggestions which WIPO is free to accept or reject. As with the Interim Report, the exigencies of a rushed schedule greatly limited the amount of time we had to review and comment on the Final Report before it was issued. Just as the Final Report therefore does not necessarily speak for every member of the advisory panel of experts, this document represents my views only. The views expressed here should not be attributed to WIPO, nor to any other member of the advisory panel unless otherwise stated at the top of this document.
  5. Background

  6. The management of Domain Names is widely agreed to be a core issue of Internet governance. Many people involved in the rapidly evolving institution-building process that will perform core functions of technical (and perhaps legal) coordination of the Internet have looked to WIPO for leadership in resolving thorny questions relating to conflicts created between global domain names and (often) regional and sectoral trademarks.
  7. In a "White Paper" formally known as the "Statement of Policy on Management of Internet Names and Addresses" (June 5, 1998) issued by the Department of Commerce of the United States of America, the United States Government called on WIPO to:

  8. "initiate a balanced and transparent process, which includes the participation of trademark holders and members of the Internet community who are not trademark holders, to (1) develop recommendations for a uniform approach to resolving trademark/domain name disputes involving cyberpiracy (as opposed to conflicts between trademark holders with legitimate competing rights), (2) recommend a process for protecting famous trademarks in the generic top level domains, and (3) evaluate the effects, based on studies conducted by independent organizations, such as the National Research Council of the National Academy of Sciences, of adding new global Top-Level Domains (gTLDs) and related dispute resolution procedures on trademark and intellectual property holders. These findings and recommendations could be submitted to the board of the new corporation for its consideration in conjunction with its development of registry and registrar policy and the creation and introduction of new gTLDs."
    The "new corporation" referred to above is now known as ICANN-the Internet Corporation for Assigned Names and Numbers.
  9. Prior to the Final Report, WIPO produced three key documents it labeled RFC1, RFC2 and the Interim Report, RFC3. RFC 3 was issued Dec. 23, 1998. WIPO also published a useful background information document and a process timetable.
  10. As an organ of the United Nations, responsible to all its member states rather than just the US, WIPO felt empowered to define its own terms of reference. In its RFC2, paragraph 12, WIPO stated that it intended to make recommendations concerning (1) dispute prevention, (2) dispute resolution, (3) process for the protection of famous and well-known marks in the gTLDs, and (4) effects on intellectual property rights of new gTLDs. WIPO thus gave itself a considerably broader and more ambitious charge than the fairly narrow one proposed by the US in the White Paper. Many, both in the trademark-owning community and outside it, objected to this ambitious agenda. The Final Report largely returns to the original mandate suggested by the U.S. Government.
  11. On April 30, 1999, WIPO issued the Final Report. WIPO's recommendations in the Final Report now go to ICANN, which can adopt them, modify them, or ignore them as it will. However, WIPO's report is likely to be influential. Few others have been focusing on the DNS/Trademark problem and politics, like nature, abhors a vacuum. Indeed, ICANN's draft registrar accreditation guidelines adopted some of the features of RFC 3, even though RFC 3 was only an interim report.
  12. A "Purloined Letter" Process?

  13. Significant parts of the WIPO process were open to public view and comment, although the actual drafting of the Report was entirely the work of the WIPO staff, operating behind closed doors. There are reasons, however, to worry about the extent to which public participation in the comment process may have been limited, or skewed.
  14. The WIPO DNS/TM process has certainly been public in a formal sense, with a series of meetings around the world, and web pages displaying documents and public comments. These web pages set out the issues at stake in the DNS/TM process and disclosed the schedule for WIPO's public meetings in various world capitals. There is no particular reason, however, to assume that anyone is necessarily going to know that those web pages are there, or would necessarily visit them. WIPO had its own small mailing list with sign-up available to people who found the web pages, but did not publicize the information about upcoming public hearings on any of the external mailing lists that would likely to draw the attention of the technical or legal communities with experience in Internet matters. Thus, for example, I noticed no postings by WIPO announcing the public meetings to any of the following Internet mailing lists of which I am a member: Cyberia-l, CNI-Copyright, Cyberprof, Cypherpunks, Domain-Policy, E-Carm, IFWP, Lawprof, Interesting People (Dave Farber), or Red Rock Eater (Phil Agre), to name only the larger and better-known lists. Furthermore, the members of the WIPO staff did not interact with the public participants on the WIPO public list other than to post the occasional announcement.
  15. There can be little debate that the public participation in the process, at least until the time I published and publicized my critique of the Interim Report, was dominated by intellectual property rights holders and their lawyers and trade associations. Furthermore, my very public intervention was hardly a design feature of the WIPO process. Members of WIPO advisory panels do not ordinarily publish reports criticizing the process of which they are a part; so far as I know, I may have been the first, and although everyone was very polite I did not get the sense it was especially welcome or that it is likely to lead to further requests from WIPO to advise it on other matters.
  16. Public participation was low for a number of reasons, including poor publicity outside the intellectual property community and especially the competition for the attention of the relatively small number of people focused on the issue of Internet governance. Most of the people concerned about Internet governance understandably focused on the debate over the structure of ICANN, which was happening at the same time, rather than on a merely advisory report that would in time become part of the ICANN process itself. Turnout at the public hearings I have attended has been small - usually under 100 and sometimes about 50, and (with the exception of the Washington DC event that followed a publicity campaign I organized) comprised almost entirely of trademark lawyers, government representatives, or Internet service providers.
  17. The relative dearth of consumer representatives, public interest groups, and citizens groups participating in the WIPO process should serve to remind us all of the many reasons why we entrust major aspects of social policy making to elected officials. Legislatures and national executives are not paragons of rectitude. But when they are elected, the same political process that may make them over-solicitous to those bearing campaign contributions imposes some form of accountability to the public at large. Equally important, presentation of a matter to an elected official is a way of putting a question onto the public agenda. In contrast, a hearing in front of a subcommittee, a vote by a house of Congress, even a publication of a proposed rule in the federal register by an unelected bureaucrat, would serve to put the public (in one country) on notice in a tolerably effective way --at least in a routine and knowable way-- of the rules that someone proposes to lay down upon them. The problem of finding a suitable means of public notice for Internet-based governance procedures is in no way limited to WIPO, and to some extent its process may have suffered the usual fate of any first-mover. Nevertheless, more could and should have been done to publicize the process to groups both on and off the Internet who were outside the core government and intellectual property communities.
  18. It is not obvious to me that all the relevant portions of governments understood what was going on, or made efforts to contact all the relevant non-government stakeholders. WIPO sent notices of its proposals to every one of its member states, but one suspects from the responses received that these were directed at the patent and trademark offices with which WIPO ordinarily corresponds. Whether these notices were then circulated to other departments is hard to ascertain; it is easy to imagine why something about "trademarks" might not get flagged as having human rights aspects, for example, as the two make an unusual pairing. Certainly in the US, for example, it appears from his testimony in Washington that the US small business advocate learned of the WIPO initiative very late in the process.
  19. The Limited Role of the Experts

  20. I joined the Experts Group during the comment period on RFC 2 and attended one of the consultative meetings in the first round. In addition, the Experts Group met in Geneva for two days in December 1998 to discuss the Interim Report that was due to issue shortly thereafter. Unfortunately, we were provided with only minimal text in advance of our meeting-some by email shortly before we left, more under the door of our hotel rooms the night before our first meeting.
  21. While our debates are, I am told, confidential, I think it breaks no confidence to say that our meeting in Geneva was not a drafting session. Rather, we were invited to comment on the issues, and discussed the texts we had been given. WIPO then revised the texts very extensively, and e-mailed us the revised versions. We had only a very short turnaround, of a few days late in the holiday season, to comment by e-mail on what was, to my eye, a wholly new document. WIPO then made some additional changes, including the insertion of new material, and on Dec. 23, 1998 WIPO issued RFC 3, the "Interim Report," which contains WIPO's first draft of its proposals.
  22. Many of the Experts attended various regional consultations in the second round. We had another two-day meeting to discuss the comments on the Interim Report in Geneva in March, 1999. Again, this was not a drafting session. We were promised time to review the text of the final report before it would be issued. In fact, chapters 2-5 arrived during what proved to the final week before publication, with the last one arriving perhaps two days before the report was issued. This made commenting in detail rather difficult. None of the Annexes were sent to the Experts prior to publication.
  23. As WIPO itself notes in the Final Report, Annex 1, this is WIPO's report, not the experts'.
  24. An Introduction to the Intellectual Property Issues

  25. The domain name system was designed to make it easy for people to remember the addresses of computers linked to the Internet. Each unique name maps to a unique IP number that is the unique identifier for a computer linked to the Internet. Under the current system, both these numbers and the associated names need to be unique. Currently only a minority of top level domains (TLDs) are open to all comers. Of the so-called generic TLDs (gTLDs), .com, .org and .net are open to anyone who can afford the small fee required to register a second-level domain. Country-code TLDs (ccTLDs) carry a suffix that identifies them with a particular nation (e.g. .fr, .uk), but they are equally accessible everywhere in the world. Of the 240 or so ccTLDs, about a quarter accept registrations from non-residents.
  26. The increasing commercialization of the Internet has focused attention on domain names. Firms, and others, increasingly seek to have an Internet presence and see securing an appropriate second-level domain name in an appropriate TLD as akin to hanging out their nameplate in cyberspace. Domain names in .com/.net/.org (and in some of the more commercial ccTLDs), however, have traditionally been issued on a first-come, first-served basis. This has caused various types of conflict.
  27. Trademarks are issued on a national and sectoral basis. With the exception of some treaty-based registration systems that allow multiple registration in a unified process, trademarks are issued by national governments-one country at a time. Further, trademarks generally are issued for one or only a few categories of goods of services at a time. Thus, a firm can trademark the word "United" for air transport, but this will not extend to moving vans unless the firm is in that business also. Trademark registrations generally require use to remain effective; while they are in effect they give the holder important rights against others who unfairly would seek to capitalize on the goodwill of the mark by confusing consumers. Equally importantly, trademarks protect consumers against those who might seek to pass off their goods as produced by the mark holder. As a general matter, however, in the US at least, trademark infringement requires commercial use by the infringer. Absent commercial use, some type of unfair competition, or a very small number of other specialized offenses (e.g. "tarnishment" of a mark by associating it with obscenity), trademark law does not make the use of the mark an offense. Thus, for example, in the United States and many other countries parody, criticism, names of pets, and references in literature, and every other use one might make of a basic dictionary word such as "united," are all permissible uses of a trademarked word. Indeed, unless the mark falls into a very small category of "famous" marks where it is considered likely that any product which bears the mark will be associated with a single source, it generally is permissible to make commercial use of a name trademarked by another so long as it is not likely to cause customer confusion. Even some types of commercial use are protected against famousness, e.g. accurate comparative advertising and news reporting and news commentary.
  28. An underlying legal issue is whether registration of a domain name that is identical to a trademarked term is in and of itself a trademark violation. Generally speaking, in the US at least, one does not violate a trademark right without commercial use (and, absent a finding that the mark is famous, likelihood of confusion). Unless, therefore, registration is itself a commercial use, mere registration without use of a domain name cannot be a violation of a trademark right. This is particularly clear in the case of trademarks in common words and in terms trademarked by more than one party. On the other hand, two courts, one in the US and one in the UK, have held otherwise. They found that a person who made a practice of registering others' trademarks for potential resale was making commercial use of those trademarks. Assuming that these decisions are correct, which is itself controversial, I do not believe that the precedents would or should apply to persons who are not in the business of registering domains that contain trademarked terms for resale on a similar scale.
  29. Types of Conflicts

  30. The Internet is notoriously international, and every one of the major TLDs gives access to sites that are accessible world-wide. [There are some "alternate" gTLDs that require pointing one's browser to a different DNS server. Although not heavily used at the moment, that may change depending on the reaction to ICANN's actions.] A system that relied on geographic distance and sectoral differentiation maps badly to a borderless world in which every participant in the global network needs a unique address.
  31. A number of conflicts have arisen between trademark holders and others who register character strings identical or similar to their trademarks. In so-called "cyber-squatting" cases, the allegation is usually that the registrant is not using the domain but rather warehousing in hopes of reselling it at a (sometimes substantial) profit. Not every string conflict, however, necessarily involves a claim of mis-use of a domain and not all warehousing is necessarily a mis-use. For example, firms sometimes acquire domains with the same name as a trademark they have registered even though they have no intention of using the domain. They do so in order to prevent someone else from using it and causing customer confusion. Similarly, firms and others sometime acquire domains for future use. A firm may register a domain name before trademarking a term as part of the often-secret process of preparing a new product or campaign. In fact, these practices have given rise to some expressed concern that without new gTLDs large amounts of the attractive part of the namespace might become unavailable to users.
  32. As part of its first round of consultations, WIPO sought testimony about the extent of the domain name/trademark conflicts. Despite this effort, in my opinion the factual record for any decisions regarding the extent to which there is a domain name/trademark problem that needs solving remains depressingly sparse. It is disappointing (if understandable given the very short amount of time available) that WIPO has not sought to do, or commission, independent quantitative research on the subject other than contacting existing registration authorities and asking for data. Nevertheless, the largely anecdotal evidence [see RFC 3, paras 254-260, for a summary of some of it] submitted to WIPO and/or available from other sources in my opinion establishes the following points with sufficient clarity to guide a modest amount of cautious future policy making pending further research:
  33. Conflicts can arise between multiple owners of a trademark in the same "string" of characters. The owners may be
    1. sectorally separate (same country, but different use or different category of goods and services), or
    2. geographically separate (same business, but different countries or regions within a country), or
    3. both sectorally and geographically separate.
  34. Conflicts can arise between trademark holders and persons with some other indisputably legitimate interest in a mark not deriving from a trademark.
  35. Domain name disputes involve a tiny fraction of the number of domains registered in the open gTLDs. The NSI policies are very favorable to trademark holders, so one would reasonably expect that a large fraction of the trademark holders whose cases fit the criteria would choose to invoke it; nothing, however, requires them to do so. The NSI dispute policy, which applies only to conflicts involving an exact string match, was invoked 838 times in 1998. This represents a 10% reduction over 1997. The decline in disputes brought to NSI is all the more dramatic when one considers that NSI registrations more than doubled in the same period from 962,000 in 1997 to 1.9 million registrations in 1998.
  36. Of course, many disputes are never reported to NSI at all because they do not fit NSI's criteria. For example, if "startrek" is a trademarked term, then attempts to register that name by non-trademark holders will be reviewable under the NSI policy. But an attempt to register "startrek1" would not be unless there were a separate trademark for that exact term. In the absence of firm data we must rely on anecdotal evidence for the extent of the overall problem. One data point is the testimony of Ms. Sarah Deutsch, Sr. Intellectual Property Counsel, Bell Atlantic, who stated that in a year her office identified 784 domain names that it considered infringed one of its trademarks, but only 10 could be resolved through NSI's dispute resolution policy. Second D.C. Consultation Transcript. Using that approximately 80:1 ratio to inflate the NSI data suggests that of the 1.9 million new domain name registrations in 1999, about 67,000 (i.e. 80*838) allegedly infringed a trademark. While this is not a trivial number, it is still only a fraction of all registrations.
  37. The suggestion that the number of domain name disputes may have reached a plateau, or even peaked, is not entirely consistent with some of the anecdotal evidence presented to WIPO, but it may be that may be that precisely those who have been most affected by the conflicts have had the most incentive to participate in the WIPO process. The Marques study, Final Report paragraph 313, also can be read to suggest otherwise, but one needs to consider whether the companies surveyed were representative or likely to be particular targets, and the very small sample size.
  38. A very small number of notorious "cyber-pirates" have been engaged in systematic registration of domain names identical to trademarked terms. Although at present there are no reliable data that permit one to estimate the percentage of "cyber-pirate" or "cyber-squatting" cases attributable to this group, trademark holders suggest it represents a considerable percentage of the problems they believe that they face.
  39. There is also an unquantified amount of what might be termed amateur domain name speculation, in which individuals who are not engaged in the wholesale registration of domains containing trademarked character strings register domains that trademark holders believe are rightfully theirs. The data on offer do not allow one to form a reliable estimate of the percentage that so-called amateur speculators make up of the total cases that trademark holders believe that they face.
  40. In many cases these individuals avoid paying for registrations by taking advantage of the 60-day period between registration and payment under the current NSI terms of service. A substantial fraction of NSI's domain name registrations are not in fact paid for at the end of the 60-day grace period.
  41. An appreciable but unquantifiable fraction of the cases alleged by trademark holders to fall in the above cases, and in particular the allegedly amateur speculator cases, in fact appear to be cases where the registrant has at least a colorable, and perhaps a very legitimate, claim to the domain name. In some cases this arises from a competing trademark, and in other cases it arises from some other legitimate commercial or non-commercial purpose, use, or competing intellectual property right or name.
  42. The overwhelming majority of the cases that have actually gone to trial in the US and elsewhere and resulted in a reported decision have resulted in victory for a trademark holder over a non-trademark holder. Every organized cyber-squatter who has been taken to court appears to have lost.
  43. There have also been some well-reported cases of attempted reverse domain hijacking in which trademark holders retracted their threat to sue the holders of domains that used the same string as their trademark. A few of these cases involved commercial, but most involved non-commercial uses of the domain name; the key elements seems to have been bad publicity for the mark holder, combined with limited likelihood of success in the U.S. courts.
  44. Trials can be expensive, and trademark holders have with some frequency found it cheaper or more expedient to offer out-of-court settlement to registrants of domain names that the trademark holder covets. Whether or not the sums involved have been large, firms managing large numbers of trademarks fear the cumulative cost of these settlements.
  45. Aggrieved trademark holders in some countries believe that their national court systems are so slow as to provide no meaningful relief for domain name registrations that they believe infringe their trademarks.
  46. Although they have no current plans to use them, some firms are warehousing domain  names corresponding to their trademarks in order to prevent competitors from registering them.
  47. The marketplace now provides a service for firms to register a name in every top-level gTLD and ccTLD which accepts such registrations through a single process. The going rate for this service is apparently less than US$10,000.
  48. Some persons have registered domain names similar but not identical to the domains held by high-profile individuals or companies. Some of these have sought to capitalize on the similarity, or on the typing errors that might unwittingly send a web browser to the site. Motives for registering these "oops" names and creating web sites appear to vary and include:
    1. Using the domains to host web sites that parody or criticize the individual (often a politician) or company;
    2. Taking advantage of the accidental traffic for relatively harmless commercial gain, e.g. to show the user an advertisement before redirecting the user to the site the user was probably looking for;.
    3. Taking advantage of the traffic for commercial gain that would arguably tarnish the reputation of the company. For example, a representative of AT&T stated that someone registered attt.com and placed pornographic materials on the site, resulting in frequent telephone calls to AT&T from consumers unaware of their typing error who wanted to know why AT&T was hosting pornography. [Currently, however, the site www.attt.com appears to be down; whois now shows that name to be registered to AT&T.]
    4. Taking advantage of poor typists who were seeking a competitor's web site.
  49. The understandable fear of trademark owners that they must assert their rights to their marks in every medium has also contributed to litigation and something of a hair-trigger approach to domain names that use the same terms as trademarks.
  50. There have been some notorious cases of "reverse domain name hijacking" in which the owner of a trademark has sought to intimidate the holder of a coveted domain name to surrender it. Examples include pokey.org, epix.com, cds.com, ajax.com, dci.com, ty.com, roadrunner.com, and veronica.org.
  51. The Final Report

  52. As in the Interim Report, WIPO's proposal depends on a chain of contracts regarding the assignment of domain names. ICANN will be at the base of the chain. It will select and contract with a small number of registries who will maintain and manage the database of domain names. The registries will in turn contract with a much larger number of registrars who will actually deal with the registrant-the customers seeking a domain name. The core of the WIPO proposal is to ask ICANN to initiate a chain of contractual obligations such that:
  53. The ICANN-registry/registrar contract would have terms requiring that all registrants agree to be bound by the WIPO takedown policy for incorrect contact details, and that all registrants agree to submit to, and be bound by the administrative dispute policy (WIPO-ADR) unless there is contrary judgment of a competent court. Final Report, paragraph 162, 196, 220(ii).
  54. Furthermore, all registries would be required to
    1. agree to adhere to decisions of the WIPO-ADR panels unless these were contradicted by a competent court, and
    2. agree not to register in any gTLD those domains identical to trademarks found as famous by the WIPO famous names panels.
  55. All registrants would
    1. agree not to sue the various dispute service providers, and
    2. agree not to sue the registrar or the registries for implementing the results of the ADR proceedings. (Final Report, para 220.) As we will see, this last agreement turns out to be significant if the registrant believes the arbitrators have erred.
  56. WIPO proposes that these rules should apply to all open gTLDs, that is to .com, .net, .org and to any and all new gTLDs that would offer registrations to anyone for sale. WIPO is not proposing that ICANN require that these rules apply to ccTLDs, although it recommends that ccTLDs adopt them spontaneously. Being contractually based, the new rules will be imposed on persons who already have contracts for domains in .com, .net., and .org as those agreements lapse. Existing registrants will be presented with the revised contractual terms on a take-it-or-leave-the-Internet basis.
  57. The Interim Report proposed that the WIPO-ADR cyber-arbitration could be invoked by any third party who felt their intellectual property rights of any sort (not just a trademark) had been affected by the domain registration. In contrast, the Final Report limits the scope of the mandatory arbitration clause to "abusive registrations" a.k.a. cybersquatting. Registrants will, however, be given the option of agreeing to more sweeping arbitration at the time of (re-)registration.
  58. As in the Interim Report, WIPO also proposes the creation of a parallel process in which a centralized tribunal would hear requests by trademark owners that their marks be certified as globally famous or well-known across a widespread geographical area and across different classes of goods and services. Marks certified in this process would be entitled to two special privileges:
  59. Registries and registrars would agree to refuse to register a domain name identical to the mark. However, this exclusion would "apply only to new open gTLDs," and "would not have any retroactive effect, i.e. if a party had registered a string as a domain name in relation to which an exclusion is later granted to another party, the first party's domain name would remain unaffected by the exclusion (but the other party could seek to obtain its cancellation through the administrative dispute-resolution procedure)." Final Report Para 276.
  60. Marks certified in the procedure would also benefit from an "evidentiary presumption" in WIPO-ADR proceedings. Any time the holder of a certified mark could "show" "(i) that a domain name was identical or misleadingly similar to the mark that is the subject of the exclusion; and (ii) that the domain name was being used in a way that was likely to damage the interests of the owner of the mark that was the subject of the exclusion. The burden of proof in the procedure would shift to the domain name registrant to justify that its registration of the domain name was in good faith and to show why that registration should not be canceled. If the domain name registrant were unable to make such a showing, the registration would be canceled." Final Report, para. 289.
  61. How one feels about the Final Report depends in significant part on what one uses as a baseline for comparison.
  62. Compared to the Interim Report, the Final Report looks good, mostly because the Interim Report was--to be blunt--both one-sided and overreaching. Although not without flaws, the Final Report is more cautious, and more balanced, than its predecessor.
  63. Compared to current practices, the Final Report also represents potential progress: With the exception of the proposed special privileges for famous marks (and assuming that some minor ambiguities can be resolved), the system proposed improves on current practices. Current practices are a combination of
    1. The precedents being built up fairly rapidly in the judicial systems of several countries; and
    2. The actions of the NSI dispute policy in what appear to be a declining number of cases per year; and
    3. The practices of registrants and trademark holders.
  64. On this baseline, the picture is generally positive. Unlike the Interim Report, the Final Report seeks to reflect and build on current law, rather than to replace it. The proposed definition of "abusive registrations" and the means proposed for adjudicating claims of abuse appear to be substantially superior to the NSI dispute policy, although some practical issues may remain to be fleshed out, particularly regarding the formal description of the policy in the Annexes, some transitional issues, and the conduct of on-line arbitrations.
  65. In one respect, however, the Final Report falls badly short. For the reasons set out in more detail below, the proposals relating to special privileges for famous marks are unjustified both in principle and in practice.
  66. Compared to my ideal, the Final Report leaves quite a bit to be desired.
    1. Privacy protections in the existing open .gTLDs (.com, .org & .net) falls to zero. Combined with procedures designed to make it much more difficult to submit false registration information, the net effect is to reduce privacy in registration and acquisition of domain names below their already low level.
    2. The Final Report is at best lukewarm in its endorsement of a privacy-enhanced global Top-Level Domain (gTLD) such as .per,.nom, or .personal.
    3. A party who loses an arbitration under the proposed procedure, especially a person living in the USA, may find it difficult to secure judicial review of the decision to take away his/her domain name.
    4. The Report assumes a level of competence in online arbitration that does not yet exist.
  67. To the extent that the Interim Report fairly captured the agenda of a segment of the Intellectual-property-owning corporate sector, the Final Report may also appear to be something of a disappointment to those interests.
  68. The (Relatively) Good News

  69. Several of the most objectionable features of the Interim Report have been removed or altered in the Final Report.
  70. Excessive scope. The proposals in the Interim Report tried to solve every imagined intellectual property problem relating to domain names, instead of concentrating on the trademark-related issues that most urgently need solution. The Interim Report went far beyond the issues raised in the U.S. Government White Paper. In so doing it threatened to create an unprecedented opportunity for trademark holders to attempt reverse domain name hijacking at the expense of unrepresented parties. WIPO instead offered a comprehensive scheme for alternate dispute resolution of almost every conceivable intellectual property dispute that might involve a domain name-including substantial non-trademark issues
  71. For example, the Interim Report would have required registrants to arbitrate claims that their domain name infringed another's "right of personality". The right of personality is a controversial doctrine by which some countries give persons, including politicians, actors, and other famous people, special rights over the use of their names. In some cases the right includes elements of privacy, reputation, protection against defamation, and even "a right of informational self-determination," i.e., a right exclusively to determine whether and to what extent others might be permitted to portray one's life story in general, or certain events from one's life. The danger was that this expansive scope would have constrained expressive activity if, for example, politicians could claim that their critics were not allowed to register the politician's names, nor perhaps even strings containing their names (e.g. contraPinochet) as domains. Indeed, in RFC 3, para 115, WIPO suggested that "any dispute concerning the domain name" could go before its proposed administrative tribunals-a phrase that apparently included claims that a domain name is libelous, offensive, guilty of lèse-majesté, contained a geographic identifier, and even contract disputes involving a domain name.
  72. In contrast, the Final Report restricts the proposed ADR procedure to cases in which the complainant alleges an "abusive registration" (cybersquatting) on the part of the registrant. The Final Report specifically disclaims covering cases other than those relating to cybersquatting on names protected by trademarks and service marks (paragraph 167). Thus, cases where both sides have valid trademarks, and claims based on trade names, geographical indications or personality rights are all explicitly excluded from the procedure. Id.
  73. One of the more common criticisms of the Interim Report was that it failed to offer a definition of cybersquatting. The Final Report offers such a definition for the first time. It defines a domain name registration as "abusive" (para. 171(1)) if,
    1. "(i) the domain name is identical or misleadingly similar to a trade or service mark in which the complainant has rights;" and
    2. "(ii) the holder of the domain name has no rights or legitimate interests in respect of the domain name;" and
    3. "(iii) the domain name has been registered and is used in bad faith."
  74. For the purposes of the third element of the definition of "abusive registration," the Final Report states in paragraph 171(2) that "the following, in particular, shall be evidence of the registration and use of a domain name in bad faith:
    1. "(a) an offer to sell, rent or otherwise transfer the domain name to the owner of the trade or service mark, or to a competitor of the owner of the trade or service mark, for valuable consideration;" or
    2. "(b) an attempt to attract, for financial gain, Internet users to the domain name holder's website or other on-line location, by creating confusion with the trade or service mark of the complainant;" or
    3. "(c) the registration of the domain name in order to prevent the owner of the trade or service mark from reflecting the mark in a corresponding domain name, provided that a pattern of such conduct has been established on the part of the domain name holder;" or
    4. "(d) the registration of the domain name in order to disrupt the business of a competitor."
  75. The Final Report emphasizes that "the behavior of innocent or good faith domain name registrants is not to be considered abusive. For example, a small business that had registered a domain name could show, through business plans, correspondence, reports, or other forms of evidence, that it had a bona fide intention to use the name in good faith. Domain name registrations that are justified by legitimate free speech rights or by legitimate non-commercial considerations would likewise not be considered to be abusive." Para. 172.
  76. Paragraph 171 is reproduced as Paragraph 15 of Annex IV, the proposed policy document. But, paragraph 172 is inexplicably absent from Annex IV. (The Annexes were not circulated to the Experts Panel prior to publication). Nevertheless, one presumes that arbitrators will understand that their primary task is to determine the good faith of a registrant in the light of this important clarification.
  77. Even in light of paragraph 172, the wording of paragraph 171 suggests a few issues that may arise:
    1. Presumably, the complainant carries the burden of proof. Although this is not explicitly stated anywhere, plaintiffs usually have the burden of proof, and in the absence of anything to the contrary, it seems right to assume it. Indeed, given that paragraph 291 suggests that the owner of mark that has been certified as globally famous is entitled to an "evidentiary presumption" that places a "burden of justifying the registration" on the registrant, no other conclusion seems possible in the ordinary case.
    2. A related question is whether plaintiff must affirmatively allege each of the three elements of an abusive registration, and in particular whether plaintiff must affirmatively allege that registrant has no legitimate rights or interests in the name at issue. This requirement seems to follow from the plaintiff having the burden of proof. Rules of court do not apply in arbitrations, but it is to be hoped that lawyers representing parties in these cases will hold their submissions to the same high standards of honesty and disclosure that would apply in a court proceeding.
    3. A slightly more difficult question is whether the complainant fully meets his burden of proof of bad faith by offering any of the types of "evidence" listed in paragraph 171(2); a considerably more difficult question is what might suffice to rebut that showing. It seems probable that in the absence of any contrary evidence, persuasive evidence conforming to any of the four categories in paragraph 171(2) would suffice to make out a case (and indeed, the door is not closed to offering some other type of evidence instead). Paragraph 171 should not be understood, however, to suggest that even a persuasive prima facie showing of this kind is conclusive; there is always room for rebuttal evidence. For example, if Plaintiff submits evidence that complainant has a valid trademark in "companyname." alleges that registrant does not, and submits evidence that registrant offered to sell companyname.com to complainant for $1000, this is sufficient evidence to make out a case of abusive registration. However, it remains open to registrant to show either that he has a valid interest in companyname, or that the offer to sell was in response to a request from the plaintiff.
    4. Although paragraph 171(2)(a) is ambiguous on the issue, it is hard to see how it could be bad faith to respond to a solicitation of a bid, especially if there had been any sort of dispute in progress between the parties and the offer to sell was part of a settlement. Any other rule would create a trap for the unwary, but given the seemingly absolute wording of paragraph 171(2)(a) the matter is not as free from doubt as one would wish. There seems little danger that understanding paragraph 171(2)(a) in the manner proposed will create a large opening for passive cybersquatting. Any party who makes a practice of registering other people's trademarks and waiting for offers to roll in is likely to be caught under paragraph 171(3), which deals with patterns of registering others' trademarks. Ideally, paragraph 171(2)(a) would be clarified to make it abusive to "initiate" an offer rather than merely "offer".
  78. The abusive registrations policy will definitely classify some name brokerages as cybersquatters. An issue is whether any domain name brokers can remain in business. A name broker who registers a coined trademark that belongs to only one firm is unlikely to be able to show that he has rights or legitimate interests in the name. Thus, anyone other than a owner or licensee of the mark who registers "coca-cola" in any gTLD for the purpose of resale (as opposed to, say, inveighing against the consumption of sugary drinks) is henceforth (and probably already) a cybersquatter. The picture is slightly murkier with regard to generic words. To the extent that generic terms such as "cars" or "fresh" or "united" cannot be trademarked, a name broker might be able to defend a claim by arguing that the complainant did not have the valid trade or service mark needed to fulfill the first prong of the definition. 
  79. One thing seems very clear. Although some, perhaps many, cases will lend themselves to easy determination on the papers, there will inevitably be a class of cases in which the arbitrators will be called upon to make credibility determinations. Suppose, for example, that registrant explains that he registered companyname.com because he calls his cat "companyname". People do call cats all sorts of strange things. Will the naked claim be sufficient? Fortunately, in the absence of any offer to sell the domain name, or the (alleged) ownership of a troupe of cats who happen to bear trademarked names, the issue will not arise. But if it does, one hopes that the arbitrators will, especially in the early cases, make clear what sort of proof they expect, e.g. a sworn affidavit, a picture of the cat on the web page, written testimony from neighbors or a vet, or whatever.
  80. 'Smorgasbord' approach to law. I criticized the Interim Report for taking what I characterized as a "smorgasbord" approach to choice of law. Critique at paragraphs 21, 155- 157. In particular, I criticized WIPO's suggestion that arbitrators should be required to refer to a set of WIPO-defined "guiding principles." In contrast, paragraph 176 of the Final Report notes that, "In applying the definition of abusive registration given above in the administrative procedure, the panel of decision-makers appointed in the procedure shall, to the extent necessary, make reference to the law or rules of law that it determines to be applicable in view of the circumstances of the case. Thus, for example, if the parties to the procedure were resident in one country, the domain name was registered through a registrar in that country and the evidence of the bad faith registration and use of the domain name related to activity in the same country, it would be appropriate for the decision-maker to refer to the law of the country concerned in applying the definition." Similarly, Paragraph 15 of Annex IV, the proposed policy document, puts the same point but in slightly less clear terms: "To the extent that the Panel makes reference to any applicable law to reach a determination, it shall apply the law or rules of law that it determines to be appropriate in light of all the relevant circumstances."
  81. What constitute "rights and legitimate interests," Final Report para. 171(1)(ii), is one area of potential ambiguity in the definition of "abusive registration." Rights and legitimate interests are creatures of law and, perhaps, custom. Solicitude for the rights of free expression of citizens differs considerably around the globe, and choice of law may be significant in a multi-jurisdictional case. The interests one can have in a name differ in other ways from country to country (e.g. the latitude given to startup companies, the nature of interests one can have in ones' own name). Which law, and which customs, apply may be a complex question if a party from Korea registers a domain with a registrar in Japan that deposits the data in a registry located in the USA, only to have a Brazilian claim that the registration infringes a trademark. It seems to me, however, that such multi-jurisdictional problems can be unraveled using traditional choice of law principles. In my view, paragraph 176 of the Final Report, which states that "In applying the definition of abusive registration ... the panel of decision-makers appointed in the procedure shall, to the extent necessary, make reference to the law or rules of law that it determines to be applicable in view of the circumstances of the case," should be understood to require nothing less. While the text of that paragraph is perhaps not as transparently clear on the issue as one might ideally wish, the rejection of any attempt to create a lex mercatoria or a free-standing trans-national law of domain names is clear from the overall thrust of the Final Report. It is doubly clear when one notes the absence in the Final Report of anything comparable to the "guiding principles" that appeared in paragraphs 198-200 of the Interim Report and that apparently might have trumped national law.
  82. Enabling censorship. Many argued that the Interim Report failed to protect fundamental free-speech interests including parody, and criticism of famous persons and corporations. So long as it is understood that these expressive activities constitute "rights and legitimate interests" in a registration - and paragraph 172 of the Final Report seems to make that clear - then the Final Report can be regarded as an unqualified improvement in this regard.
  83. Potential for intimidation. The Interim Report created an expensive loser-pays arbitration process with uncertain rules that I argued would intimidate persons who have registered into surrendering valid registrations thus enabling increased "reverse domain name hijacking". The Final Report substantially solves this problem by providing for much clearer rules, at least outside the context of famous names. The loser may still have to pay in many cases (the arbitrators get to decide who pays), but since a finding of "bad faith" is required before the registrant pays this seems acceptable. It may be more unfair to make loser pay in a proceeding involving a famous name, however, since the complainant gets to have an "evidentiary presumption" that they win even if the name is merely "misleading similar" (whatever that means) - and it seems invidious to presume bad faith by means of an evidentiary presumption in cases where the similarity is an issue.
  84. Relied on potentially unenforceable contracts. Because the contracts of adhesion proposed by WIPO were so one-sided, there was reason to believe they would not be enforceable in court and that the entire proposal was therefore unworkable as the contracts were arguably unconscionable, and doubly adhesive. The Final Report solves the first part of this problem by making the contract much fairer. The equities are now in favor of enforcement, or at least not against it, and in my opinion the contract cannot therefore be called unconscionable. However, contracts which come as part of a standard form that is not subject to bargaining are called "contracts of adhesion"; online contracts of adhesion are enforceable in some parts of the U.S., see, e.g., Hill v. Gateway 2000, Inc., 105 F.3d 1147 (7th Cir. 1997), cert. denied, 118 S. Ct. 47 (1997), and there are (somewhat controversial) proposals to make this the uniform rule. The standard policy argument in favor of enforcing adhesive contracts is that consumers have a choice in the marketplace and can always switch suppliers; if it happens that all the suppliers use the same term, this is considered to be evidence that the term is efficient or, at least, that there is insufficient consumer demand for an alternate term. This policy justification does not fit well when the party offering the adhesive contract, the registrar, is doing so because of an adhesive contract it signed with the registry. If the justification for enforcing adhesive contracts turns on the fundamentally competitive nature of the market, and the idea of sovereign consumers choosing among alternatives, then it would not be appropriate to allow the adhesive contract proposed by WIPO to shelter behind this rationale, as it is proposed that ICANN require that all gTLD registries/registrars to force the identical terms from all their customers. Thus, unlike the ordinary contract of adhesion, the terms will neither be bargained-for (since the registrar cannot bargain on this issue, being contractually required to use the arbitration clause), nor even potentially subject to competition. The good news here, perhaps, is that so long as the arbitrators do their jobs properly there will be few if any parties willing to mount the legal challenge. Furthermore, any party that attempts such a challenge risks arriving in court with the equities against them.
  85. Matters for ICANN

  86. The Final Report essentially leaves two important matters to ICANN's sound discretion. While all should welcome WIPO's conclusion that there are no insurmountable intellectual property barriers to the creation of privacy-enhanced gTLD and to new differentiated and even undifferentiated gTLDs in general, some comments about strategy and timing seem appropriate.
  87. New gTLDs

  88. WIPO concludes that, "on condition that the proposed improved practices for domain name registrations, the proposed administrative dispute-resolution procedure and the proposed measures for the protection of famous and well-known marks and for the suppression of abusive registrations of domain names are all adopted, new gTLDs can be introduced, provided that they are introduced in a slow and controlled manner which takes account of the efficacy of the proposed new practices and procedures in reducing existing problems." Final Report, paragraph 343. This conclusion is welcome and even enlightened compared to some of the views that one hears. I think, however, it understates the urgent need for one or more privacy-enhanced gTLDs. In general, it seems plausible that a "controlled manner" can contribute to minimizing disputes in the new gTLDs. Controlled, however, does not necessarily mean "slow"-rather, it means "orderly." Indeed, I believe that one of most important things that ICANN could do to reduce speculation in domain names, and thus reduce the conflicts that are most likely to harm trademark owners, is to clearly state an upgrade path for the introduction of new gTLDs as soon as possible.
  89. The issue of new gTLD's provides a classic example of the economics of scarcity, of network effects, and the problem of path dependence. Competition for attractive names in the .com, and to a lesser extent .net and .org domains is driven by a combination of the following factors:
    1. Second-Level Scarcity: There currently are only three open gTLDs, causing a scarcity of available common English words, a perception of an over-all shortage, and a hoarding mentality. This accounts for the gold-rush approach to registrations.
    2. Network effect: To a significant extent, .com has achieved a small priority of place in the English-speaking Internet-using world as the TLD of choice for commercial undertakings (although, interestingly, it does not appear to have this priority in the non-English-speaking world). If there were a substantial number of competitors for .com, and it became widely understood that any Internet address was substantially arbitrary, then the understandable fears of trademark owners that consumers might naturally gravitate to .com in search of their product will be greatly alleviated.
    3. TLD Scarcity. If TLDs were managed on free-market principles, the pent-up demand for additional attractive second-level domains would surely have resulted in the creation of many new gTLDs already. The move by some ccTLDs to brand themselves as de-facto gTLDs (e.g. .tm, .md, .io, .to) demonstrates the existence of a market.
  90. Furthermore, any proposal for creating a small number of new gTLDs, slowly, that does not make it clear that many more will be created in a predictable fashion according to a fairly predictable schedule risks repeating the scarcity problem at least, and perhaps the network effect problem as well. This is the problem of path-dependence. Perhaps counter-intuitively, opening only a small number of domains in some ways may be more harmful to trademark owners in the long run than opening so many that the market for domains becomes flooded (and the value of cyberpiracy/cybersquatting becomes greatly reduced as the number of equally attractive TLDs grows). A world with many, differentiated, gTLDs will help further educate users to the reality that the presence of a particular word in a domain name does not necessarily map to a company whose product may share that name. A substantial increase in the number of gTLDs would also make it easier to provide attractive second-level names to each of several parties with an interest in the same character string (e.g., multiple holders of sectoral or national trademark interests in the same name).
  91. A fair solution requires some creativity. Although there may be some technical constraints on the rate at which new TLDs safely can be proliferated, and also on the absolute number that can comfortably be established under the current architecture, I understand that there is no technical reason why at least a hundred new TLDs could not be created over a year at a rate of, say, two per week. While I do not necessarily advocate that many, one should not assume that a solution with many gTLDs will be worse than one with very few and the attendant scarcity mentality that shortages breed. Indeed, Jon Postel proposed creating up to 150 more TLDs. See Final Report, para. 305.
  92. In my opinion, the issue of gTLDs open to all comers should not be separated from the very closely related issue of restricted, or structured/chartered, gTLDs. A restricted gTLD is a domain with a defined and limited purpose. There have been a plethora of proposals for restricted gTLDs, including suggestions of .biz (for firms only), .nom (for personal names only), .per (ditto), .museum, .law, .fun, .xxx (a domain which might make content filtering more easy for those who choose that option), and many others.
  93. One can imagine at least two kinds of restricted gTLD:
    1. An open restricted gTLD, where the purpose of the domain is announced but registrations are taken in the same manner as .com, i.e. first-come-first-served, with no policing of bona fides of registrant nor of intended use prior to registration. What policing exists is driven by third-party complaints after the fact; and both registry and registrar seek to avoid participation in those disputes. This was the original design of .net (for ISPs and related providers), and .org (for non-profits); it is fair to say that as regards those domains the structure has broken down.
    2. A closed (or moderated) restricted gTLD, where not only is the purpose of the domain announced, but the task of approving registrations by policing the bona fides of registrant and/or the intended use prior to registration is delegated to an Approval Authority. The Approval Authority for each moderated gTLD would be an appropriate body, with relevant expertise regarding the purposes of that domain. Policing would exist would in the first instance be the gate-keeper function of the Approval Authority, but there could also be scope for third party complaints to the Approval Authority. The Authority would have the authority to de-list registrants or transfer registrations if it upheld complaints. There might also be scope for meta-policing via complaints (to ICANN?) that the Approval Authority was doing a bad job.
  94. Specialized gTLDs should be created, with a mixture of open and moderated structures. A variety of procedures for helping ICANN to select the new gTLDs to be created might be used. In addition to top-down planning of the sort advocated above, some gTLDs might be allocated according to popular demand, expressed and measured in some fashion akin to the procedures used by the Usenet Volunteer Vote-Takers, or the procedures currently being developed for membership in ICANN. In at least some of these there would be no need for the sort of administrative dispute procedures advocated by WIPO because the uses were purely non-commercial (with, perhaps, some sort of policing or challenge mechanism to prevent abuse) or because access was carefully controlled by a responsible body.
  95. Zero Privacy

  96. The Interim Report provided zero privacy protections for the name, address and phone number of individual registrants. Under current practices, a registrant can achieve privacy by giving inaccurate contact details, but this will become riskier under the plan proposed in the Final Report since inaccurate contact details will be grounds for rescinding a registration.
  97. It is widely albeit not universally agreed that commercial users do not have a strong claim to privacy-enhanced domain names. Many, however, believe as I do that individual registrants who are not engaged in a profit-making endeavor do have a right to keep their identity and contact information private. One day everyone on the planet may have their own domain. Data collection and publication requirements suited to businesses are not appropriate for ordinary people who register a domain and who understandably do not wish their name, telephone number, and address published on the world wide web. It is even less suited to social, ethnic, religious, and political groups who have reason to fear retaliation if the information were disclosed. Every collector and keeper of this personal data should be held to the highest standards of protecting individual privacy.
  98. Given that it will become more difficult to camouflage identity in the DNS, the absolute minimum that is required to ensure that the emerging Internet architecture retains a modicum of personal privacy is to create a privacy-enhanced zone on the Internet. There needs to be a domain where individuals can register for non-commercial purposes without having their personal contact details visible for the world to see. Unlisted/unpublished telephone numbers are commonly available in many countries, and it is difficult to see what harm a similar system would cause in the DNS.
  99. That said, I accept that WIPO is correct that a privacy-enhanced non-commercial gTLD cannot be implemented without at least some further discussion. There remain some issues to be decided, notably what precisely constitutes the "non-commercial" use of a domain name. (Is a religious group's solicitation of funds non-commercial? The sale of sacred candles? And so on.) My difference with WIPO on this subject is as much one of tone as substance, as I have a very different sense of the urgency with which the remaining issues need to be tackled.
  100. Some will say, with some justice, that all this is too kind, that the absence of any privacy protections in the existing gTLDs is a scandal whose ill effects should not be replicated, much less exacerbated by the WIPO proposals. There are two replies to this critique. First, privacy as such has not traditionally been seen as an intellectual property issue, but rather as a matter of human rights. As such, the WIPO process was not the place to look for improvements in this area. That is ICANN's responsibility and ICANN's duty. The WIPO report represents merely a sort of lower bound, not a statement of optima. It remains open to ICANN to convene a process that discusses the human rights issues implicated by its work; indeed, there are issues in the world even more important than the protection of intellectual property. (Furthermore, to the extent that one believes in rights of personality, one's interest in one's personal data may in fact be an intellectual property right.) Second, one should not forget that the DNS, like every electronic data collection and dissemination project, remains subject to various national laws. The trend is towards increasing protection for personal data, and it may be that these laws and directives are the arena in which privacy rights ultimately will be vindicated.
  101. The Bad News

  102. The most significant bad news concerns the treatment of famous marks. Certain other matters, particularly the procedural details also raise substantial concerns in their current form.
  103. Treatment of Famous Trademarks

  104. The Interim Report proposed new, cumbersome, unwarranted, and potentially limitless procedures to protect "famous" trademarks. These proposals are repeated essentially unchanged in the Final Report, although some of their malign effects are reduced due to the more limited scope of the proposed ADR system. Both the ad hoc procedure for certifying which marks are sufficiently globally famous to qualify for the special advantages WIPO proposes and, the consequences of achieving that status are unsatisfactory at both theoretical and practical levels.
  105. WIPO proposes that a centralized tribunal be impaneled that would rule on ex parte applications by trademark owners that their mark was sufficiently famous to be declared globally famous. This entire project described in Chapter Four of the Final Report is at best premature. While many nations have, consistent with their treaty obligations and national law, developed standards for identifying nationally famous and/or sectorally well-known marks there is no consensus procedure for identifying globally famous or globally well-known marks. Furthermore, any mark which is globally famous or well-known will also be nationally famous or well-known in most nations. As such it already benefits from existing, substantial, protections under national law and is therefore already quite well protected, by the courts-and by the WIPO-ADR procedures when they follow national law.
  106. World-wide there probably are already at least tens of thousands of identified nationally or regionally famous and well-known marks. As the procedures for identifying these marks become more routinized and widespread there are potentially hundreds of thousands, or even millions, on a world-wide basis. WIPO does not (as indeed it should not) propose to give special gTLD protection to each of the many nationally famous and well-known marks, but rather to create a new category of globally famous and well-known marks. However, it appears that no one-WIPO included-has any idea how many marks this would be. In the Interim Report WIPO suggested that only "a small number of names is involved...it is likely that famous and well-known marks that may qualify ... number in the hundreds, rather than the thousands." RFC 3, para. 216. Interestingly, no estimate is provided in the Final Report.
  107. The job of trying to find appropriate and definite criteria is a thankless one, as the inability of the Committee to come up with something better demonstrates. WIPO rightly rejected mechanistic criteria such as the number of countries in which a mark is registered, as these too are manipulable, and they also fail to measure whether a mark has the global fame of exceedingly common names such as "Coca-cola." Unfortunately, rather than recognize that the task is perhaps impossible, but certainly not yet possible, WIPO pressed on.
  108. There are several conceptual problems with WIPO's proposal:
    1. The "exclusion" protection for certified famous marks provides marks owners with rights not currently found in law, notably the right to block non-infringing uses of the name they have trademarked;
    2. The criteria WIPO proposes be used for identifying which marks are sufficiently famous to enjoy enhanced protection are vague, and in one appears to require evidence whose prejudicial effect will outweigh any probative value (as WIPO argues ADR is needed because cybersquatting affects many trademark holders, making being the subject of cybersquatting an indication of global famousness seems especially odd.);
    3. No one has any idea how many marks will qualify. Not only is there no upper limit, WIPO has now retreated from its low estimate in the Interim Report;
    4. The proposed "evidentiary presumption" applies far too broadly, and is in any case an inappropriate remedy.
  109. Problem: Provides rights not found at law. The protections proposed for famous marks exceed those currently available at law, thus violating the fundamental aim stated in the Final Report that "the goal of this WIPO Process is not to create new rights of intellectual property, nor to accord greater protection to intellectual property in cyberspace than that which exists elsewhere." Final Report, paragraph 34. As WIPO itself notes, "the provisions of the Paris Conventions and the TRIPS Agreement are directed at the protection of famous and well-known marks against the registration or use of other infringing marks. Domain names, of course, are not the same thing as marks and are used for many purposes other than the identification of a producer or seller of goods or services." Final Report, paragraph 258. WIPO is to be commended for its honesty in noting that these international treaties do not create a right to prevent registration of domain names. This, however, leaves the question of where that right comes from.
  110. Once a mark is identified as famous, the proposed remedies and benefits are over-broad, as they block non-commercial uses of the same name, and hard to understand (and, it appears, over-broad), as applied to the multiplicity of similar names.
  111. WIPO proposes two types of benefits to globally famous marks. The "exclusion" would automatically block the registration of a domain identical to the trademark in any new gTLD. This exclusion fails to reflect the fact that the international consensus on the protection of famous and well-known marks extends only to protection against commercial use of those marks. Whatever protection these marks are entitled to under treaty extends only to protection against others making commercial use of the mark.
  112. Similarly, although the laws of many nations provide protection against dilution of a trademark this too can require commercial use, albeit not necessarily a showing of likely customer confusion. For example, the U.S. federal Trademark Dilution Act specifically excludes non-commercial use of a mark from its coverage. 15 U.S.C. § 1125(c)(4)(B). Thus, in Lockheed Martin Corporation v. Network Solutions, the court stated,

  113. "in Panavision and Intermatic, the fact that the defendant's conduct impeded plaintiff's use of its trademark as a domain name was not the determining factor in finding that the defendant's use was diluting. If impeding use of the trademark as a domain name were the only factor, the court in Panavision would not have asserted that registration of a trademark 'as a domain name, without more, is not a commercial use of the trademark and therefore not within the prohibitions of the Act.' Panavision, 945 F. Supp. at 1303. All prior domain name registrations corresponding to words in a trademark impede the trademark owner's use of the same words for use as a domain name. The Internet, however, is not exclusively a medium of commerce. The non-commercial use of a domain name that impedes a trademark owner's use of that domain name does not constitute dilution."
    985 F. Supp. 949, 959-60 (C.D. Cal. 1997). As the court noted, "Internet users may also have a free speech interest in non-infringing uses of domain names that are similar or identical to trademarks." Id. at 964 n. 9.176. Since a registration of even a famous mark is not, without some commercial use, necessarily infringing, it is unreasonable to give even globally famous marks a blank pre-emptive right blocking registrations in all gTLDs
  114. WIPO makes two arguments for extending the protection of famous marks beyond what is currently available in law. The first argument need not detain us long. Famous marks, WIPO argues, deserve something extra in the DNS because they have extra rights in national law, "The administrative procedure is, however, rightly available to all and does not give expression to the separate international protection that already exists for famous and well-known marks", Final Report, paragraph 262; an exclusion mechanism is justified because "it seems correct in principle that famous and well-known marks are recognized in international law as being subject to special protection," id., paragraph 263. Rather than "extend such protection," the exclusion will "give expression to it in the DNS". Since it is already conceded that this extra protection does not derive from the Paris Convention or TRIPS, it can only derive from national law; as such the appropriate level of protection will be applied when judges and arbitrators decide specific cases. In any event, WIPO's argument misses the point. The issue is not how to replicate the relative degrees of protection given to different types of trademarks. The issue is how to craft processes that provide trademark owners who have legal rights that are cheap to violate in the DNS, but expensive to vindicate in the courts, with a cheap, fast, fair means of enforcing their rights-without going overboard and violating some other rights in the process. The WIPO-ADR already does this.
  115. WIPO's second argument is more practical than legal: "We also consider that it could be highly economically wasteful, in view of the experience in the existing open gTLDs over the past five years, to add new open gTLDs without any safeguard against the grabbing or the squatting of famous and well-known marks by unauthorized parties in those new open gTLDs." This argument is not persuasive, either on its own terms, or in light of WIPO's consistently-repeated promise that its objective was to replicate existing legal relations, not expand the rights of intellectual property owners. It is not the law, in the US at least, that one commits an actionable wrong against even a famous trademark by making many non-commercial use of it. Thus, an exclusion which works to prevent critics of a corporation from registering its name in a new gTLD for the purpose of critiquing it gives that company an advantage not justified by law (and thus violating WIPO's own terms of reference for itself).
  116. That said, there is clearly a good argument for some transitional rule to cover the first micro-seconds or days when certain new gTLDs are opened up. Such a rule would address the competing interests of the multiple possible trademark holders as well as non-trademark holders with a legitimate interest in a given domain name. But the need for a rule to prevent (or mediate) a short-lived stampede should not be leveraged into a rule that creates a permanent preference of unlimited scope. Nor is it clear why the same rule would apply to, say, a non-commercial gTLD.
  117. WIPO itself has had an international working group seeking to identify criteria for world-famous marks for several years. See Final Report, para. 284. This group has failed to produce either a list of world-famous marks or a set of definitive criteria for identifying such a mark. See id. As WIPO accepts the general principle that "the goal of this WIPO Process is not to create new rights of intellectual property, nor to accord greater protection to intellectual property in cyberspace than that which exists elsewhere," Final Report, paragraph 34, it seems most in keeping with this principle to wait until the bodies already charged with finding definitive criteria for all purposes do so.
  118. Problem: No Upper Limit The ad hoc procedures WIPO proposes to create have no structural incentive for keeping the number of specially privileged marks within reasonable bounds. Put simply, the participants in WIPO's proposed process have no incentive to say "no" to anyone claiming that his mark is famous or well-known. The consequence for the DNS could be a substantial erosion of the available namespace as an increasing number of names, including common words, are reserved and made unavailable in every gTLD. WIPO responds that the "best safeguard" against this fear is "discipline and rigor in relation to the criteria for assessment of entitlement to an exclusion," Final Report, paragraph 274, but it offers no explanation of how this is to be achieved.
  119. If it cannot scrap the entire idea, ICANN needs to set an upper bound for the number of marks that could benefit from the process. Once such a ceiling was in place, the decision-makers would have an incentive to be cautious about whether marks are sufficiently famous to qualify, for fear that over-liberality in the initial decisions would leave out more deserving marks whose owners were slower to request certification. An initial ceiling could be 500, 1000, or even more, trademarks. Since any initial number is arbitrary, the ceiling could be subject to review (up or down) by ICANN or some other neutral body every five or ten years.
  120. WIPO explains its opposition to an upper limit in terms that reveal just how necessary a limit will be: "We consider that a quota could operate in an entirely arbitrary manner. The selection of the level of the quota would, for a start, be arbitrary. The level could work arbitrarily against marks which become suddenly famous, whose owners might be prejudiced by the previous filling of the quota." Final Report, paragraph 269. With these words, WIPO admits that there is no quota it can propose that would be large enough, and that whatever the quota is agreed, WIPO fears that the process it proposes would fill it. This alone should be conclusive evidence that the proposal for certifying globally famous names is fatally flawed. (Recall that in the Interim Report WIPO suggested only "hundreds" of trademarks would qualify; now it makes no estimate.)
  121. The claim that a quota would be arbitrary is persuasive only if the quota is set so low that it gets used up. If it is the case that WIPO's criteria will be applied with 'discipline and rigor' then it should be possible to find a quota high enough not to cause arbitrary results and yet low enough to reassure the Internet community that the mechanism will not be abused.. The suggestion that a quota would work to unfairly disadvantage a suddenly famous mark like Viagra that was a late arrival on the scene, see Final Report paragraph 269, can be met either by setting the right quota, or by setting a quota with a warning zone. Suppose, for example, that the quota were set at 1000 trademarks, with the understanding that if the panels actually allocate the entire quota, they can have another 100 so long as the criteria for granting certifications of sufficient famousness were officially revised to make granting certification more difficult. In this scenario there would never be a moment in which there were literally not more certifications available. The disadvantages would be that were the criteria ever tightened this would in effect give a windfall to successful early applicants who would not have met the tightened criteria. As a result a "flexible" ceiling system (and indeed, a low and inflexible ceiling also) might lead to a flood of applications at an early stage. Encouraging early applications may not be all bad, however, as it will give the decision-makers a better sense of the marks against which they should weigh each applicant.
  122. If the entire proposal for special privileges is not abandoned, ICANN could, and should, in any event review the quota every few years.
  123. Problem: Vague & Prejudicial Criteria. The fear that 'discipline and rigor' would be short-lived at best is strengthened not only by WIPO's inability to make an estimate of the total number of marks likely to qualify, but also by the vague and in one case prejudicial criteria that WIPO proposes be used to rule on applications for global famousness. To begin with, WIPO offers the six vague and manipulable considerations for determining if a mark is famous or well-known issued by WIPO Standing Committee on Trademarks, Industrial Designs and Geographical Indications and set out in Para. 282 of the Final Report.
  124. WIPO then added a seventh "non-exhaustive" criterion not approved by the committee: "Evidence of the mark being the subject of attempts by non-authorized third parties to register the same or confusingly similar names as domain names." Final Report, Paragraph. 283. No justification was offered for this suggestion in the Interim Report other than it would serve "to accommodate the specificities of the protection of famous and well-known marks in relation to domain names" and, again, no further justification for this seventh factor is offered in the Final Report. Why precisely a mark should be more likely to be considered globally famous because it happens to attract the attention of a single enthusiastic domain name speculator, numerous parody sites, or nettlesome critics, is unclear. Indeed, if a firm is attracting the attention of many critics who register names similar to it as a form of protest, this seems to be the weakest case for special protection. Indeed, as WIPO argues ADR is needed because cybersquatting affects so many trademark holders, making being the subject of cybersquatting an indication of global famousness seems especially odd.
  125. Furthermore, the experience of the United States with the federal Anti-Dilution Act suggests that when faced with a blameless trademark holder, a predatory domain name registrant, and a set of rules that allows the decision maker to do equity, even federal judges cannot resist declaring that trademarks one never heard of are famous. Adding the seventh criterion ensures that the decks will be stacked before the special tribunals as well, with predictable results.
  126. Problem: Inappropriate Additional Remedies and Benefits-"Evidentiary Presumption". WIPO also proposes that the owners of certified globally famous trademarks benefit from an "evidentiary presumption". The "evidentiary presumption" is in fact a shifting of the burden of proof. The mechanism is not explained as clearly as one would like, but it appears that after having a mark certified as globally famous, a mark-holder would begin an ADR proceeding against a registrant in the ordinary manner. Instead of having to allege and carry the burden of proof that the registrant had no right to use the domain name, and the registrant was acting in bad faith, the extra-famous mark holder would only have to persuade the arbitrators that the domain is "misleading similar" to his mark and that it is being used in a manner which damages his interests. Upon that lesser showing (and apparently without the benefit of knowing how the arbitrators have ruled on the complainant's suggestion that the burden of proof be reversed), the registrant will then have the burden of proving "justification" for her use of the domain name. Final Report, Paragraph 289.
  127. There are four problems with this idea. First, the term "misleadingly similar" is vague and likely to be interpreted in an over-broad manner. Second, the suggestion that any use of a domain name that might "damage the interests of the owner of the mark" sweeps far too broad, and will invite harassment of many legitimate domain name registrants. Third, reversing the burden of proof-requiring someone to prove their good faith-is inherently unfair. Fourth, the procedure proposed for the ADRs in Annex V of the Final Report is designed in a manner that fails to ensure that the respondent will given sufficient notice as to the case which he has to answer.
  128. What is "misleadingly similar"? WIPO's Final Report does not define the term "misleading similar". While one could profitably attach some sort of definition to the name (it sound as if it has some relationship to the US idea of "likelihood of confusion" except that in the context of a famous mark one would expect more a dilution than a confusion test), thus perhaps lessening the ill effects of this proposal, I think it a fundamentally unprofitable exercise because the entire idea is founded on a misconception. In the US, and I suspect elsewhere as well, it is not seriously disputed that a domain name is not a trademark. Therefore, what determines whether the registrant of a domain name is infringing the rights of any mark-holder, whether or not the mark is famous, is how the domain name is being used. The issue therefore is not whether the domain is the same as, or close to, or very close to a trademark, but whether the name plus the uses made of that name dilute the mark.
  129. What will "damage the interests of the owner of the mark"? The suggestion that the owner of a certified famous mark need only allege that the registrant's use of the domain name somehow damages the "interests" of the owner of the mark requires too gentle a showing and risks inviting frivolous and harassing claims. If Bob is running a site parodying Alice's product, or a site that seeks to gather potential litigants to sue for an alleged product defect in Alice's product, or an "anti-Alice" site with a confusing name such as A1ice, or a site in which Bob's protest group is coordinating a boycott against Alice for her labor or environmental policies, it is quite likely that Bob is "damaging" Alice's "interests"--yet each of these uses of a domain name similar to or even identical to Alice's trademark is legal, and some are commendable. At the very least, and as a deterrent to harassing applications, Alice, the applicant, should be required to affirmatively allege that she believes there is no reasonable case to be made that Bob is engaged in non-commercial activities.
  130. Is burden-shifting fair and appropriate? The suggestion that an person should be considered a presumptive cybersquatter merely because they register a domain name similar to a well-known trademark is a novel one. One rarely if ever finds cases in the law of civilized countries where a person must prove their innocence of an offence by a preponderance of the evidence. This may seem a trivial point, but one must recall that the critical issue is likely to be whether the registrant has"rights and legitimate interests" in the name. The remaining uncertainty about what law will be applied to determine this question, and precisely what the terms mean, is worrying. Combined with the likelihood that many respondents will be unrepresented consumers with no experience in arbitration, much less the on-line variety, while the famous mark holders are likely to the parties with the greatest experience at both and the finest representation, and the issue of the burden of proof becomes more significant.
  131. WIPO argues that since the ADR is now more limited than in the Interim Report, this evidentiary presumption can no longer be used in a way that is likely to damage recognized rights of free speech. Alas, WIPO's sanguine description of the likely uses of its proposal likely will prove optimistic. Furthermore, WIPO suggests at several places in the Final Report that the ADR procedure proposed is only a first careful step towards the more ambitious plan it outlined in the Interim Report. One should thus think carefully about the potential long-term consequences of Chapter 4 in this context.
  132. The nature of available judicial review

  133. As in the Interim Report, challengers to domain names registrations who lose at the ADR stage may appeal to any competent court (including the jurisdiction where the registrar is located). And, as in the Interim Report, a party who loses a WIPO-ADR could face very substantial difficulties in securing judicial review of the decision. Challengers who lose an arbitration can bring a de novo action in any court with jurisdiction over the registrant, but in many cases the registrant will have only an inconvenient forum, and sometimes no forum at all, if he loses the ADR proceeding. The Final Report does not solve this problem. Instead it ameliorates it by reducing the volume of cases that potentially are covered under the policy. While welcome, this is not an actual solution: As further explained below, in so doing this raises question of which court and which law will be available after the WIPO-ADR is concluded. WIPO's suggestions that registrants who improperly lose their domain names will have a judicial remedy does not address the nature of the cause of action (if any) that may be available to a registrant who has lost his domain, especially if the registrant was not a trademark holder. Indeed, in some cases there may be no way to bring suit against the successful complainant. The only recourse will be to sue the registry itself, a right that WIPO suggests should be waived in the domain name registration contract.
  134. A complainant who wishes avoid the WIPO-ADR can bring the action in any court that has jurisdiction over the registrant. Suppose that Alice, the complainant, lives in New York, and Bob, the registrant, lives in Prague. If Alice can persuade a New York court to assert jurisdiction over Bob because he is using the domain in an infringing manner with effects in New York, then she can bring suit where she lives. On the other hand if Bob has merely registered the domain but made no infringing use of it, Alice probably must go to Prague to bring the action. (Alice may have waived her ability to in rem complaint against the registry once in enforces the ADR decision by virtue of the terms proposed for her contract in paragraph 220(iii) of the Final Report.) The WIPO-ADR offers Alice a potentially attractive means of avoiding the expense and uncertainty of hiring foreign counsel and risking the vagaries of a foreign legal system. If Alice loses, and she wishes to bring suit anyway, she has the same options she had before the WIPO-ADR.
  135. Suppose, however, that the WIPO-ADR rules that Bob, the registrant, should surrender his domain name but that Bob wishes to challenge the outcome, perhaps because he believes that under Czech law he has a valid right to the name that the arbitrators failed to recognize. Under the WIPO policy the decision goes into effect within seven days, Final Report paragraph 216, so Bob has a week to find a court with jurisdiction over the Alice to hear his request for a stay. Without the WIPO-ADR Bob probably would have defended the action in a court in Prague, giving him the benefits we traditionally accord defendants, and especially defendants who are ordinary individuals and small businesses: a convenient venue, familiar law, local language and local counsel, and the choice of law principles of the local court. Instead, unless Alice has sufficient contacts with Prague for the court there to assert jurisdiction over her, now Bob must shoulder the burden of being the plaintiff in a New York court, with potentially unfamiliar and more expensive procedures, a different local language, and what to Bob will be foreign counsel. The New York court may use different choice of law and different substantive principles than the Czech court. And, Bob will now be the plaintiff instead of a defendant and must shoulder the burden of proof. Indeed, if Bob seeks injunctive relief to prevent the WIPO-ADR decision from going into effect immediately, Bob will have to shoulder a heavy burden of proof indeed.
  136. A significant fraction of these problems could be avoided if, as a condition of allowing Alice to invoke the WIPO-ADR procedures, Alice was required to consent to in advance to jurisdiction in Prague, the jurisdiction where the Bob resides, in the event that Bob wished to stay or to overturn the WIPO-ADR decision. The consent could be limited so as to exclude nations that are parties to neither of the relevant international agreements on intellectual property, e.g. the Paris Convention for the Protection of Industrial Property or bound by the Agreement on Trade-Related Aspects of Intellectual Property Rights (TRIPS Agreement).
  137. As noted above, WIPO recommends that, as a condition of being allowed to (re)register a domain name, a registrant be required to waive his ability to bring suit against either the registry or the registrar once any ADR decision has been enforced against him. See Final Report paragraph 220(iii). As a result, the Final Report creates a perhaps unintentional incentive for registrants to commence litigation before the ADR result is announced. This may work to defeat the purpose of enterprise to the extent that the purpose is the swift and inexpensive resolution of disputes, although it also forces the registrant to become the plaintiff thus taking on the burdens of finding a forum and carrying the burden of proof.
  138. As noted in my critique of the Interim Report, there may be a more fundamental problem also. WIPO has not been able to meet its own design goal that "the right to litigate a domain name dispute should be preserved," Final Report, paragraph 148, and "the availability of the administrative procedure should not preclude resort to court litigation by a party. In particular, a party should be free to initiate litigation by filing a claim in a competent national court instead of initiating the administrative procedure, if this is the preferred course of action, and should be able to seek a de novo review of a dispute that has been the subject of the administrative procedure." Final Report, 150(iii). The question remains: what exactly does Bob tell the court if he loses the WIPO-ADR? Recall that the "administrative" decision goes into effect in seven days and remains in effect until countermanded by a competent court.  Although some nations provide procedures for the review of "arbitration" it is unclear if this "administrative" procedure would necessarily qualify. If it is an "arbitration" then some cause of action other than an action to review the arbitration is required.
  139. The difficulty of finding such a cause of action is illustrated by US law, which does not provide for general review of arbitrations. The US Federal Arbitration Act does not provide a means of review, since that act limits the court's review to

  140. "(1) Where the award was procured by corruption, fraud, or undue means.
    "(2) Where there was evident partiality or corruption in the arbitrators, or either of them.
    "(3) Where the arbitrators were guilty of misconduct in refusing to postpone the hearing, upon sufficient cause shown, or in refusing to hear evidence pertinent and material to the controversy; or of any other misbehavior by which the rights of any party have been prejudiced.
    "(4) Where the arbitrators exceeded their powers, or so imperfectly executed them that a mutual, final, and definite award upon the subject matter submitted was not made."
    9 U.S.C. § 10.
    None of these four factors will ordinarily apply.
  141. U.S. Courts do not ordinarily review "administrative" decisions of  private parties (as opposed to government agencies), unless there is some claim of tort, breach of contract, or violation of some other legal right. Having lost the domain name, Bob must now frame a cause of action, relying on a legal right, that will get a court's attention. Bob thinks the arbitrators improperly ignored a legal basis for his claim of right to the name, or incorrectly decided that Bob was a liar.
  142. Several possible-sounding claims are probably hopeless:
    1. Bob might be able to frame some sort of claim of "tortious interference with contractual relations" against Alice, based on his contract with his registrar, but that seems a poor bet when he specifically agreed to the ADR procedure in his contract.
    2. If Bob is a non-commercial user and there is no claim of bad faith or fraud on the part of the arbitrator. Bob will not be able to claim a violation of his right of free expression because the damage was caused by a private party, not the government. He has little actual damages, and it is in any case unclear who has been negligent or behaved tortiously. There is no statutory right at issue; Bob's right to an injunctive relief requires a claim of right under law and it is not at all clear where this would be found.
    3. Under RFC 3, there is no contract between Bob and Alice for the court to adjudicate, and Bob has no claim against Alice under his contract with the registry. Even if there were a contract between Bob and Alice entered into at the start of the WIPO-ADR, I am unsure that they could by contract manufacture a cause of action that a U.S. court would accept gave Bob standing. (Other jurisdictions have procedures for judicial review of arbitration, but that is not how the U.S. arbitration act works.)
    4. Nor, absent a trademark of his own, is Bob likely to have a claim against Alice under Alice's subsequent contract with her registry.
    5. Incidentally, if Bob has a trademark identical to his domain name, and the arbitrators just ignored it for some strange reason, he can claim that Alice is violating his trademark. But the strength of that claim will turn in substantial part on how Alice is using the mark, not on what Bob was doing which would have been the subject of the case but for the WIPO-ADR. It is not difficult to imagine a case where the two parties are not in fact infringing each other, and a court applying national law would have found for Bob if he were the defendant. But as Alice is no more guilty of trademark infringement under the relevant national law than is Bob, she will win the court case-and keep the domain Bob would have had but for the WIPO-ADR.
  143. A slightly less hopeless argument would be "tortious interference with a prospective business advantage" against Alice. The Restatement (Second) of Torts § 766B, states, "One who intentionally and improperly interferes with another's prospective contractual relation (except a contract to marry) is subject to liability to the other for the pecuniary harm resulting from loss of the benefits of the relation, whether the interference consists of (a) inducing or otherwise causing a third person not to enter into or continue the prospective relation or (b) preventing the other from acquiring or continuing the prospective relation." However, US courts have frequently imposed more stringent limiting conditions on this tort than the Restatement (Second) formulation might suggest. For example, "In order successfully to assert a claim of tortious interference with prospective business advantage, plaintiff must show that 'the defendant's interference with business relations existing between the plaintiff and a third party, either with the sole purpose of harming the plaintiff or by means that are dishonest, unfair, or in any other way improper.' PPX Enters., Inc. v. Audiofidelity Enters., Inc., 818 F.2d 266, 269 (2d Cir.1987)." Similarly, "In order to state a claim for tortious interference with prospective economic advantage, a plaintiff must show (1) business relations with a third party; (2) defendants' interference with those business relations; (3) defendants acted with the sole purpose of harming the plaintiff or used dishonest, unfair, or improper means; and (4) injury to the relationship." Purgess v. Sharrock, 33 F.3d 134, 141 (2nd Cir.1994). This will usually be difficult to prove, or even to allege in good faith: any Alice will be able to argue convincingly that harm to Bob was not the sole purpose of the ADR in that Alice sincerely wanted the domain name for herself. Furthermore, assuming that it was the arbitrators who erred, and there was no fraud by Alice, Bob cannot in good faith claim that she used "dishonest, unfair, or improper means" to win the ADR.
  144. If the above analysis is correct, then for many-perhaps most or all-registrants who lose an ADR, their dispossession is the whole of the law. On the other hand, challengers who lose an ADR lose nothing more than their costs as they retain whatever right to litigate they may have had previously.
  145. Other Issues

  146. A few other issues deserve mention, some relating to material added in the Final Report and thus not exposed to public comment.
  147. Nature of ADRs. The entire WIPO edifice is predicated on the ability of dispute resolution services to provide relatively quick and inexpensive on-line arbitrations. If the arbitrations are not quick, trademark owners will be unhappy. As the arbitrations become more expensive, the potential for abuse of registrants grows as they will be increasingly wary of an expensive loser-pays system. Indeed, at some point the increased expense will make the system unattractive to all parties.
  148. It bears noting that the world has almost no experience with online arbitration. It has even less with international online arbitration. Similarly, there is considerable experience with commercial arbitration, and even with domestic business-to-consumer arbitration, but no significant experience with international business-to-consumer arbitration. Implementing this system will thus take us deep into uncharted waters. Arbitrators are going to have to make credibility determinations based on paper records, generated by people who may in many cases be arguing in a language other than their native tongue. Standards of proof are not clearly spelled out, and these will matter.
  149. The novelty of the ADR procedure is not necessarily a reason to avoid it, although it explains why many fear it. If nothing else, it requires that ICANN proceed cautiously, and that it schedule the procedure for regular and frequent review.
  150. Transitional Issues. The Final Report glosses over substantial transitional issues regarding consent to jurisdiction. Making the whole world agree to suit where the registrar is located, see Final Report, paragraphs 111 & 147(ii), will be fair once there are lots of registrars. But not until then. Even after there are many registrars distributed over the world, the rules for changing registrars will be particularly significant in this context, since the registrar one selects will affect the location in which one can be sued.
  151. If, as seems likely, we swiftly move to a world where there are many gTLD registrars distributed around the world, then it will be increasingly less problematic to ask registrants to agree to jurisdiction in the country where the registrar they deal with is located--although I wonder whether the consumer laws of some countries will allow this in a consumer sales contract if it means agreeing to foreign jurisdiction. Until that point, however, even to require an agreement to be sued where the registrar is located is to require contractual agreement to a trial in foreign forum that may be far away, in a foreign language, with foreign procedures.
  152. Required Representations. WIPO proposes that every registrant in a gTLD be required to make "a representation that, to the best of the applicant's knowledge and belief, neither the registration of the domain name nor the manner in which it is to be directly or indirectly used infringes the intellectual property rights of another party." Final Report, paragraph 109(i). This requirement is inappropriate and over-broad.
  153. One might argue that this requirement is pointless since it is not at all clear whether a mere registration of a domain name can, without use, infringe the intellectual property of another. Indeed, in the US, a person cannot violate the Lanham Act or the Dilution Act without commercial use of another's mark. The requirement may seem even more innocuous when one considers that many lawyers I have spoken to take the view that in all but the most obvious cases of unfair competition the only way to know whether a proposed trademark, much less a domain name registration, is an infringement is to see what a competent court says. Whatever one may think of this type of legal realism, it demonstrates that many lawyers will have no fear advising people to register whatever they want and let the courts sort it out later; the rule will therefore stop only the highly scrupulous (and those without legal advice) without in any way daunting the ethically dubious.
  154. I am concerned, nonetheless, that this requires too broad a representation by the registrant. If Alice and Bob both have registered the same trademark in different sectors and/or jurisdictions, and Alice is aware of Bob's registration, how can Alice give a representation in good faith that her registration, which prevents Bob from registering his mark in that gTLD, does not "indirectly" infringe "the intellectual property rights of another party."? By registering the trademark in a gTLD Alice indirectly interferes with Bob's rights, as but for Alice he could register the domain.
  155. Furthermore If Bob is ignorant of Alice's registration (the proposal creates no duty to investigate), he can make the representation, but there seems no sound reason to reward Bob's ignorance at the expense of an Alice too honest to make the representation.
  156. Proposed Procedures. None of the procedures proposed in Annex V have had any public comment. They were not sent to the Panel of Experts for review or comment prior to publication. [Indeed, none of the Annexes were shown to the Panel of Experts prior to public issuance.] The procedural proposals thus should be treated as the very rough draft they are. They cannot be adopted in their current form without working severe injustice.
  157. The proposed procedures contained in Annex V impose several unfair time limits on registrants and invite various forms of abuse.
  158. First, the date of commencement of the proceedings is the date that Complaint is received by the Arbitration Service Provider, not the date that the registrant has actual (or even constructive) notice of the complaint. The idea that time begins to run before a defendant is notified of a complaint against him violates all established notions of due process in the civilized world. Admittedly, there are difficult issues regarding what constitutes sufficient notice in an online-proceeding. Given the different ways in which e-mail is used, however, WIPO however proposes the most unfair rule available.
  159. WIPO's proposed rule could easily be abused.
    1. The proposed rules contemplate e-mailed notice by the Provider to the registrant;
    2. Time starts to run when this notice is e-mailed, not when it is read: "a notice or other communication shall be deemed to have been received on the day it is delivered or, in the case of telecommunications or Internet modalities, transmitted," Final Report, Annex V, Art. 3;
    3. Nothing in the rules requires that the complainant make any effort to contact the registrant prior to filing the request for arbitration, nor is there a requirement that such contact be alleged by the complainant. A registrant thus may have no reason at all to expect to be subject to an arbitration, and will not be on notice that they should check their mail.
    4. Thus, a person who fails to check his email for ten days can lose by default.
  160. I presume this is a drafting error, but it is a serious one. It cannot be WIPO's intention to have a more solicitous procedure for canceling a registration due to unreliable contact details (Final Report, paragraphs 122,-123), than for taking away the domain name of a person who may be reachable, acting in good faith, and away from email for a week and half. As an initial matter, the rules must be re-written to require evidence of receipt of a letter before action, and the submission of evidence that such a letter was received or served should be a condition precedent for the commencement of the arbitration. If such evidence cannot be offered the complaint should be dismissed, and the complainant directed to the takedown procedures for allegedly unreliable contact details.
  161. The ten day period is in any case too short.  A complainant gets as much time as he wishes to prepare a complaint. In the ten days allotted to the registrant he must not only receive the notice, but prepare his entire defense. For a person who may be an unrepresented consumer, with no familiarity with the relevant arbitral or legal rules, this is not a very long time. And in that period he must
    1. Decide whether to seek representation;
    2. Write and submit his sole statement in his defense, Annex V, Art. 8;
    3. Collect and submit any relevant documents and a schedule of such documents, Annex V, Art. 8; and
    4. Have the defense, and possibly the documents translated into the language of the arbitration, which will ordinarily be the language of the registration agreement, Annex V, Art. 22.
  162. Ten days (minus the time it takes to get actual notice!) is simply inadequate for this, especially in the absence of any warning that the arbitration is imminent.
  163. Furthermore, in cases where the complainant benefits from a certified globally famous mark, there appears to be no provision for putting the registrant on notice as to the burden of proof he may have to shoulder. If there is a dispute as to whether the domain name is "misleadingly similar" (and given that the term is not defined, such disputes are inevitable), it seems only fair to require the arbitrators to first decide whether they accept this allegation (after hearing from the registrant), in order to then put the registrant on notice as to the burden of proof he faces.
  164. Five Proposals for ICANN

  165. The special privileges for famous marks Chapter Four should be abandoned. The proposal to set up a tribunal to determine in an ad hoc and potentially biased fashion whether certain marks rate special privileges on the Internet is fundamentally mistaken. It is mistaken because:
    1. It is unnecessary: the ADR system proposed in chapter 2 should provide fully adequate protection for famous or well-known marks.
    2. It is premature: there is a process in place to determine the criteria by which globally famous and well known marks should be identified. It has yet to come up with specific criteria.
    3. It is biased: In addition to relying on the vague criteria in existence, WIPO has added a seventh criterion of its own whose sole effect will be to prejudice decision-making.
    4. It is potentially unlimited: In the Interim Report, WIPO suggested that hundreds rather than thousands of marks would qualify for special privileges. That estimate never seemed credible. Now, WIPO chooses to make no estimate. Furthermore, WIPO in effect admits that is unable to suggest any upper limit to the number of marks that might qualify.
    5. It gives privileges beyond what the law would provide:
      1. The remedy of an exclusion is broader than would exist in law (at least in the US) since it extends to legally valid uses of the trademarked term.
      2. The remedy of reversing the burden of proof in an ADR means that some people will be presumptive cybersquatters - which is also contrary to how the law works.
  166. All of chapter 4 should be scrapped. If it cannot be scrapped, then some serious tinkering is required:
    1. A ceiling on the number of marks must be imposed
    2. A fair process must be put in place to select the institution that will administer the tribunals (maybe it should be WIPO, but this needs discussion).
    3. Much more thought needs to given as to what flows from a determination that a mark is globally famous. In particular the procedural consequences of reversing the burden of proof much more careful consideration than they appear to have received to date. Until this can be done, that benefit cannot reasonably become part of the package.
  167. Clarify the burdens and defenses in an ADR. For the avoidance of doubt, and keeping in mind the potentially international nature of the arbitrations, a few things should be made clearer in paragraph 171, and the cognate sections of Annexes IV and V:
    1. Exactly what needs to be alleged and proved, and the standard of proof;
    2. That political comment, parody, consumer complaints, and other similar expressive activities are a legitimate use of a domain name. Also, more thought needs to be given to the position of a start-up company: since it is common in Internet-related businesses to register a name early in the process, and the business may not have a trademark or many of the traditional indicia of permanence at that stage.
  168. At an absolute minimum, the text of paragraph 172 or something very much like it should be inserted into the formal policy adopted by ICANN. Paragraph 172 is inexplicably absent from Annex IV, the proposed policy document. Without this critical clarification, the policy is potentially inimical to freedom of expression.
  169. Similarly, the text of Paragraph 15 of Annex IV at least should be harmonized with paragraph 176 of the Final Report. Ideally, the final text would make it clearer than do either of those paragraphs that the arbitrators have a duty to figure out what law applies and then to use it. The alternative approach, which invites the arbitrators to make up their own rules, not only leaves an openings for decisions by the arbitrators inconsistent with the relevant national law, but also would create greater incentives for defendants to rush to court as soon as an ADR began.
  170. Establish a privacy-enhanced gTLD as a high priority. The quid pro quo for worsening the effective privacy available in the existing gTLDs should be to establish a non-commercial privacy-enhanced gTLD as soon as possible.
  171. Fix the glitches. Annex V needs extensive work. The most serious drafting errors concerns the procedural timetable. The procedural document will need to be revised to make the timetable fair, and in particular to ensure that registrants have actual notice of the ADR, and an appropriate interval to prepare their defense. Public comment may reveal other problems.
  172. There are also some substantial issues regarding the transitional period to a world of many registrars, especially regarding consent to jurisdiction. Similarly, the extent to which the required consent is fair will depend on the ease with which existing NSI customers can change registrars. It should be recognized that none of the Annexes have been subjected to the same level of non-WIPO review as the text, and that large parts of the text (including the proposed definition of cybersquatting) are themselves new and have never been exposed to public comment.
  173. The way forward. Whatever ICANN does, it should recognize two things:
    1. Several key parts of the Final Report are new or substantially different from the Interim Report. In particular, the definition of cybersquatting is wholly new, and the Annexes are so new they were never even shown to the advisory Experts group. As a result, key concepts and suggestions have yet to be exposed to public comment and debate.
    2. Much of the WIPO proposal requires a leap into relatively uncharted territory. For example we have almost no experience with the type of online international business-to-business and especially business-to-consumer arbitration proposed in the Final Report. As a result, ICANN's ultimate adoption of any of these proposals should be hedged with some type of "sunset" clause which would force a regular review of the outcome of its decisions.

About the author

A. Michael Froomkin is a Professor at the University of Miami School of Law in Coral Gables, Florida, specializing in Internet Law and Administrative Law.

Professor Froomkin is a Foreign Associate of the Royal Institute of International Affairs and a Fellow of the Cyberspace Law Institute. He is on the Editorial Board of Information, Communication & Society and Lex Electronica (Cybernews), and a Consultant Editor of Amicus Curia, the journal of the Institute of Advanced Studies (London, England). He is also on the Advisory Boards of BNA Electronic Information Policy & Law Report, the Cyberlaw Abstracts of the Legal Scholarship Network, the Privacy Exchange, and the Journal of Online Law.

Before entering teaching, Professor Froomkin practiced international arbitration law in the London office of Wilmer, Cutler & Pickering. He clerked for Judge Stephen F. Williams of the U.S. Court of Appeals, D.C. Circuit, and Chief Judge John F. Grady of the U.S. District Court, Northern District of Illinois. Professor Froomkin received his J.D. from Yale Law School, where he served as Articles Editor of both the Yale Law Journal and the Yale Journal of International Law. He has an M.Phil in History of International Relations from Cambridge University in England, which he obtained while on a Mellon Fellowship. His B.A. from Yale was in Economics and History, summa cum laude, phi beta kappa, with Distinction in History.

Professor Froomkin is a U.S. citizen. He is married, and has two children. His publications are available on the world-wide-web at http://www.law.tm .

Revision History

1.0a corrected typo in paragraphs 71, 154(2).