I. Introduction: Claims in Search of a Remedy
One of the fallouts of disruptive inventions is the need for new laws to counter their unexpected consequences. As it concerned the Internet, these consequences included a new tort of registering domain names identical or confusingly similar to trademarks and service marks with the intention of taking unlawful advantage of rights owners. Prior to 2000 the only civil remedy for “cybersquatting” or “cyber piracy” was expensive and time-consuming plenary actions in courts of competent jurisdiction under national trademark laws. A stab at providing a simpler and quicker alternative for alleged cybersquatting had been implemented in 1995 by the then sole registry/registrar of domain names, Network Solutions Inc. (NSI).
NSI’s solution, which was heavily criticized by both rights holders and domain name registrants, was to provide mark owners with a form of injunctive relief by suspending the domain name, arguably at the expense of domain name holders who were deprived of their “property” without due process. However, because only a court could determine the ultimate question of rights, NSI’s limited solution garnered dismal ratings by rights holders. Before 2000, there had been a handful of Lanham Act cases in the United States, for trademark infringement, dilution, and unfair competition, in which courts began grappling with the new tort, distinguishing it from trademark infringement and beginning the task of identifying applicable principles and factors that would justify divesting registrants of their domain names.
These disputes and the rising incidence of cybesquatting energized governments and intellectual property interests — working through the World Intellectual Property Organization (WIPO) in 1998-1999 — to cooperate in establishing a mechanism for combatting the new tort. WIPO’s Final Report, published in April 1999 (WIPO Final Report) sets out in great detail a series of recommendations for a supra-national online arbitral mechanism. These recommendations were quickly transformed by the Internet Corporation for Assigned Names and Numbers’ (ICANN) (created in 1998) into the Uniform Domain Name Dispute Resolution Policy (UDRP), which it implemented in October 1999. Upon implementation of the UDRP, the NSI Policy instantly became history. Also in 1999, Congress enacted an amendment to the Lanham Act, the Anticybersquatting Consumer Protection Act (ACPA).
It is important to underscore that an online dispute resolution proceeding under the UDRP is not a court case even though rights holders are granted affirmative relief against adjudicated cybersquatters. The UDRP is an even more powerful tool because domain name registrations are cancelled or transferred to rights holders without court intervention or a right to appeal beyond the right to contest awards in courts of competent jurisdiction, which in the United States would be an action under the ACPA.
This article focuses on the emergence and development of a jurisprudence of domain names under the UDRP; point out some of the core principles; and discuss the factors applied in distinguishing lawful from unlawful registration of domain names corresponding to trademarks.
II. Basic Ingredients of the Arbitral Process
Although rights holders were privileged in being granted a speedy and cost-effective means of adjudicating their domain name disputes, the WIPO Final Report put them on notice that their rights were no greater than what is granted under statutory law. It stated that the purpose of the proposed arbitral process “was not to create new rights of intellectual property, nor to accord greater protection to intellectual property in cyberspace than that which exists elsewhere.” Rather,
the goal is to give proper and adequate expression to the existing, multilaterally agreed standards of intellectual property protection in the context of the new, multijurisdictional and vitally important medium of the Internet…
Not only that, but it was
not intended that the means of according proper and adequate protection to agreed standards of intellectual property should result in a diminution in, or otherwise adversely affect, the enjoyment of other agreed rights…
And, finally, the scope of the procedure was limited so that it was available
only in respect of deliberate, bad faith, abusive, domain name registrations or “cybersquatting” and [was] not applicable to disputes between parties with competing rights acting in good faith.
There were two fundamental drivers for the emergence of a domain name jurisprudence. The first was the policy consensus described above: “[no] new rights [were created]”; the proposed arbitral regime was “not intended … [to] result in a diminution in, or otherwise adversely affect, the enjoyment of other agreed rights”; and it was “available only in respect to deliberate, bad faith, abusive, domain name registrations.”
The second driver, also established in the WIPO Final Report, was that “[t]he decisions taken under the procedure would be made available publicly.” ICANN expressly directs providers servicing the arbitrations to publish decisions. This official requirement enables parties and panelists to access an accumulating database of reasoned decisions, not unlike the databases of common and statutory law decisions parties rely on in framing arguments in the courts. I will return to this in a moment, but the point to be emphasized is that the availability of easily accessible databases of rulings encourages reliance on past decisions as precedent, and as the reliance grows, so does the jurisprudence.
When it came to transforming the WIPO proposals into the UDRP, the contending stakeholders submitted further comments aimed at influencing ICANN’s final language in their favor: trademark owners “suggested that the definition should be expanded to include cases of either registration or use in bad faith, rather than both registration and use in bad faith,” while individual and non-commercial interests “suggested changes in language that would narrow the scope of the definition of abusive registrations.” They “sought to restrict the scope of the examples of bad-faith practices in paragraph 4(b) and . . . to expand the scope of the ‘legitimate use’ safe harbors in paragraph 4(c).”
ICANN Staff rejected both positions. To trademark owners, it pointed out that [the] WIPO report, the DNSO recommendation, and the registrars-group recommendation “all required both registration and use in bad faith before the streamlined procedure would be invoked.” In rejecting the individual and non-commercial interest, ICANN Staff explained that “[e]ven if none of the three circumstances [in paragraph 4(c) were] present, the administrative procedure would still not apply to a dispute where the domain -name holder can show its activities are otherwise legitimate.”
Both WIPO in its Final Report and ICANN in its Second Staff Report (which implemented the UDRP) believed that decision-making under the proposed arbitral process “should lead to the construction of a body of consistent principles that may provide guidance for the future.” Whether or not “the construction of a body of consistent principles” was initially thought capable of leading to a jurisprudence, there has, in fact, emerged a “body of consistent principles [that is] provid[ing] guidance” for balancing the rights of disputants involved in cybersquatting claims. Emblematic of this emergence is WIPO’s publication of an Overview of the law as it has developed, now in its third edition (2017) which it has appropriately denominated a “Jurisprudential Overview.”
Implicitly, the WIPO consensus accepted a proposition endorsed by ICANN that the later produced Jurisprudential Overview accepts expressly, namely, that domain names can be identical or confusingly similar to trademarks yet lawfully registered as long as they are not targeting complainants’ marks.
III. Establishing Metes and Bounds of Parties’ Rights
As part of the implementation process for the UDRP, ICANN entered into agreements authorizing service providers to appoint panelists to hear and decide complaints of cybersquatting (or “abusive registration of domain names” in WIPO’s terminology). Until panelists began filing their decisions, there was no detailed body of law addressing the issue of abusive registration apart from court decisions adjudicating trademark disputes. Panelists (principally drawn from litigation and trademark bars) were invited to establish one.
Panelists are authorized under Rule 15(a) to “decide a complaint on the basis not only of the statements and documents submitted and in accordance with the Policy [and] these Rules” but also of “any rules and principles law that [they] deem applicable” (emphasis added). They also had for guidance the WIPO Final Report, which emphasized that the proposed arbitral regime was not intended to suppress legitimate competition or to restrain commerce.
Panelists began with the tablets of general principles handed down from ICANN in the form of a minimalist Policy and set of procedural Rules the meaning of which they were expected to unpack and elaborate upon in reasoned, publicly accessible decisions. It should surprise no one that once the process of construction got underway, other panelists began accepting, rejecting, adopting, refining, modifying, and citing as authority what they received and in so doing identified the principles, fleshed out the evidentiary demands, and compiled the factors for proving or rebutting cybersquatting.
There is never certainty that when a new legal process is set in motion that it will result in the emergence of a set of consistent principles and thence into a jurisprudence. To be successful, the process had to be at once predictable and consistent. The hope was that “with experience and time, confidence will be built up in the credibility and consistency of decisions made under the procedure so that the parties would resort less and less to litigation.” This is precisely what has happened. In the words of one Panel, the principles laid down by earlier Panels and the factors applied in assessing parties’ rights are “worthy of some deference.”
Whether “some deference” comes under the rubric of consensus or precedent is not without controversy. Another panelist believes that “despite the undoubted value of prior decisions, it should not be forgotten that panelists are bound by the UDRP and the Rules made under the UDRP which requires a panel to make its decision in accordance with ‘the Policy, these Rules and any rules and principles of law that it deems applicable.'” This view notwithstanding, consensus and precedent have merged into a single concept. It is rare for Panels not to cite earlier decisions supporting their reasoning.
The point can be illustrated by examining five decisions filed within the first year of the UDRP. This is not to suggest that the jurisprudence stopped growing. Later decisions could as equally be cited for clarifying principles already identified, accepting them with tweaks, or creating new core principles to address different factual circumstances.
The first decided case in January 2000 involved <worldwrestlingfederation.com>. The respondent defaulted but had contacted the complainant by e-mail three days after registering the domain name and “notified complainant of the registration and stated that his primary purpose in registering the domain name was to sell, rent or otherwise transfer it to complainant for a valuable consideration in excess of respondent’s out-of-pocket expenses.” His offer to sell the domain name to the rights holder — the first of the four circumstances of bad faith under paragraph 4(b) of the Policy — is a classic example of bad faith registration. However, as the Panel further noted, it was “clear from the legislative history that ICANN intended that the complainant must establish not only bad faith registration, but also bad faith use.” Thus the follow-up question: if the domain name is passively held, can there be bad faith use? The Panel’s not entirely satisfactory answer was that the respondent “‘used’ the domain name as [that term is] defined in the Policy.”
A better answer came a month later in the second decided case (2000-03) involving <telstra.org>. The Panel explained that passive holding of a domain name can support a finding of abusive registration because “the concept of a domain name ‘being used in bad faith’ is not limited to positive action; inaction is within the concept.” It explained that “[o]ccupying an entry in the DNS is ‘use’ . . . [because] it has a blocking function.” While mere failure to maintain an active website does not automatically result in a finding of bad faith use, the stronger the mark the less credibility a respondent has in claiming good faith. (The reverse is also true: the weaker the mark the more persuasive must be the evidence for cybersquatting). The Panel concluded that “[g]iven the Complainant’s numerous trademark registrations for, and its wide reputation in, the word <TELSTRA> … it is not possible to conceive of a plausible circumstance in which the Respondent could legitimately use the domain name <telstra.org>” (emphasis added).
The third case, also decided March, involved <telaxis.com> and <telaxis.net>. Since the domain name registration predated the trademark there could not, by definition, have been a registration in bad faith. In other words, rights holders of a post-acquired mark have no actionable claim for cybersquatting. At best the dispute involves “the competing rights and legitimate interests of two parties in the domain names.” The Panel added that “[g]iven the nature of this dispute it is properly solved by … by litigation in a forum of competent jurisdiction.”
The fourth case, number 16 to be decided (also March 2000) involved a dictionary word domain name, <allocation.com>. The Panel determined that even if a domain name is identical to a trademark, the complainant still has to prove it was registered and is being used in bad faith. The Allocation case is particularly important in establishing investor rights to their domain names. Respondent-resellers prevail when they acquire domain names for their semantic rather than their trademark values. As the Panel aptly noted:
The difficulty lies in the fact that the domain name allocation.com, although descriptive or generic in relation to certain services or goods, may be a valid trademark for others. This difficulty is expounded by the fact that, while ‘Allocation’ may be considered a common word in English speaking countries, this may not be the case in other countries, such as Germany.
Nevertheless, the complainant failed to prove bad-faith registration. The Panel held that the complainant failed to demonstrate the respondent “at the time of registration of the domain name allocation.com knew or should have known of the existence of the German trademark Allocation” and that there was “no evidence suggesting that the domain name allocation.com ha[d] been chosen by Respondent with the intent to profit or otherwise abuse Complainant’s trademark rights.”
In the fifth case, involving the descriptive phrase “smart design” (number 993 to be decided, October 2000), the Panel found the complainant at fault for overreaching its statutory rights and sanctioned it for “reverse domain name hijacking” (RDNH). The question, in this case, was whether renewal of registration amounted to a new registration; if it did it could then be argued that the domain name postdated the mark. However, the facts supported the respondent’s showing that it had rights or legitimate interests in the domain name and thus had not acted in bad faith. The Panel issued a scathing assessment of the complainant’s claim that has been cited in many subsequent RDNH decisions:
The Panel is unable to assess the Complainant’s state of mind when the Complaint was launched, but in the view of the Panel the Complaint should never have been launched. Had the Complainant sat back and reflected upon what it was proposing to argue, it would have seen that its claims could not conceivably succeed. Even assuming that its potpourri of constructive and quasi-constructive bad faith arguments were valid, they all start from the renewal, the renewal being treated for these purposes as a re-registration.
These decisions established the following core principles: (1) the complainant must prove its claim with evidence that the respondent both registered and is using the domain name in bad faith; (2) such proof can be direct, circumstantial, or inferential based on the totality of evidence offered; (3) passive holding of well-known or famous marks is inferentially bad faith unless rebutted with a persuasive explanation; (4) marks composed of common terms and descriptive phrases demand more persuasive evidence from the complainant that the respondent it them “in mind” when registering the domain name; (5) rights holders of marks acquired after registration of corresponding domain names have no actionable claims under the UDRP; (6) renewal of registration is a continuation of registration rather than a new event that restarts the clock for measuring bad faith; and (7) overreaching rights will incur the sanction of reverse domain name highjacking.
So established are these core principles that it is unusual for panelists to depart from them.
IV. Emergence of a “Common Law” Jurisprudence
As these five early cases illustrate, UDRP jurisprudence emerged incrementally through acceptance and refinement of the initial core principles, flexible enough to be applied to both commonly encountered as well as new factual circumstances. A month after the Allocation case was decided, for example, in a case involving <eautoparts.com> the Panel found that
The weakness of the EAUTO trademark makes it difficult for Complainant to argue that Respondent lacks a legitimate interest in the domain name eautoparts.com. That is because this domain name eautoparts.com is descriptive of a business that offers, through the Internet, information about or sales of automobile parts, and it is inappropriate to give Complainant a wide monopoly over all domain names, even descriptive ones, that incorporate the mark EAUTO.
Over the years, this line of reasoning has been followed in numerous decisions. Eauto, LLC has been cited as authority in dozens of cases (as have, incidentally, the Telstra and Allocation cases).
Similarly, in holding complainants accountable for abuse of the UDRP, once a clearly articulated analysis was presented, RDNH became a more common response to complainant overreaching. A three-member Panel in case number 2000-1151 (2001) cited Smart Design approvingly as well as noting an earlier case in the no sanction was imposed because “the Policy was so new.”
Nevertheless, nine years after the Panel in first held that complainants had the burden of proving conjunctive bad faith, it had second thoughts. In renouncing its earlier construction of the Policy, the Panel reasoned that his colleagues “seem to have largely overlooked the language of the Policy regarding the respondent’s representations and warranties.” The Policy creators (the Panel believed) never intended a result that allows registrants who have registered disputed domain names in good faith to take advantage of and prosper from complainants’ emergent reputations.
Instead of the traditional approach interpreting the representations provision (Paragraph 2 of the UDRP) as applying to the date of the registration of the domain name, the new construction proposed imposing on respondent a continuing obligation of lawful use. The Panel rested its new construction on the proposition that for certain kinds of opportunism, registrants should forfeit their registrations for disputed domain names regardless of whether they were registered in good faith. The theory is referred to as “retroactive bad faith registration.”
Retroactive bad faith was immediately attacked as undermining predictability and consistency, and (although it took several years of decisions) it has become a dead end. The concern for preserving these twin principles was expressed by many Panels even before retroactive bad faith became an issue. In this regard, two decisions were important markers in the history of the jurisprudence: Time, Inc. (2001) and PAA Laboratories (2004)
In Time, the majority held that a decision “should consist of more than, ‘[i]t depends [on] what panelist you draw.'” In PAA Laboratories, the Panel (sympathetic to the concept of retroactive bad faith even before it was announced but resistant to accepting it) stated that “the Panel wishes to clarify that its decision under this element is based on the need for consistency and comity in domain name dispute ‘jurisprudence’. Were it not for the persuasive force of the cited decisions, this Panel would have expressed the view that paragraph 2 of the Policy demonstrates that references to “registration” in the Policy were probably intended to be references to ‘registration or renewal of registration.'”
As a later Panel put it in 2016: “If a consensus developed that a line of prior decisions had reached the wrong result, and if panels generally adopted a new approach on an issue, this Panel also would be open to considering whether a new approach was appropriate, both substantively under the Policy and in order to promote consistency.” The Panel’s emphasis on consistency is a significant factor in its arguing against changing current law, which construes “‘evidence of the registration and use of a domain name in bad faith’ . . . merely [as] an evidentiary presumption which may be rebutted on a full consideration of all the circumstances of the case.”
The cases cited above illustrate that the emergence and growth of a domain name jurisprudence came about, and is still evolving, in a manner similar to the common law through a deliberative process that includes parties and representatives in their pleadings and arguments (to which only decision-makers have access) and that is most clearly apparent in Panel decisions as they approve, comment on, criticize, or cite earlier decisions for their own conclusions.
As domain name jurisprudence has grown in depth and complexity, it has subsumed the Policy and Rules. What this means in a practical sense is that parties and their representatives must first look at the jurisprudence to understand the current state of the law and only secondarily at the language of the Policy and Rules. It is the ongoing construction of the Policy and Rules that ultimately settles whether a party’s submission succeeds or fails in a UDRP proceeding.
 Network Solutions’ Domain Name Dispute Policy Statement Revision 02 (last modified Sept. 9, 1996. The Policy is no longer available online but a copy is attached as an Appendix to Steven A. McAuley, The Federal Government Giveth and Taketh Away: How NSI’s Domain Name Dispute Policy (Revision 02) Usurps a Domain Name Owner’s Fifth Amendment Procedural Due Process, 15 J. Marshall J. Computer & Info. L. 547 (1997).
 Its shortcomings are described in Carl Oppedahl, Analysis and Suggestions Regarding NSI Domain Name Trademark Dispute Policy, Vol. 7, Issue 1, Article 7 (1996); and the Steven A. McAuley article, supra note 1.
 Hasbro, Inc. v. Internet Entertainment Group, Ltd., 40 U.S.P.Q. 2d 1479 (W.D. Wash. 1996) (adult entertainment site at domain name candyland.com is tarnishment of CANDYLAND trademark for children’s games; motion for preliminary injunction granted); Intermatic Inc. v. Toeppen, 947 F. Supp. 1227 (N.D. Ill. 1996) (The court characterized the defendant as a “spoiler” who prevented the trademark holder from doing business on the Internet under its trademark name unless it paid the Respondent’s fee); Panavision International v. Toeppen, 141 F.3d 1316, 46 U.S.P.Q. 2d 1511 (9th Cir. 1998) (“Toeppen’s intention to arbitrage the ‘intermatic.com’ domain name constitutes a commercial use”)
 The Management of Internet Names and Addresses: Intellectual Property Issues, Final Report of the World Intellectual Property Organization Internet Domain Name Process (April 30, 1999) (hereafter, “Final Report”).
 ICANN Second Staff Report on Implementation Documents for the Uniform Dispute Resolution Policy, Paragraph 4.1(c): The Current Question Is the Form of the Implementation Documents, not the Nature of the Policy Itself. In Santiago, the Board adopted a Uniform Dispute Resolution Policy based on the final report of WIPO, the consensus recommendation of the DNSO, the recommendation of the group of approximately twenty registrars that had prepared implementation documents for a voluntary policy, and a significant number of written and oral comments.
 15 U.S.C. §1125(d) (Cyberpiracy Prevention).
 Id., Paragraph 34.
 WIPO Final Report, Paragraph 153.
 Id., Paragraph 149(v).
WIPO Overview of WIPO Panel Views on Selected UDRP Questions, Third Edition (“WIPO Jurisprudential Overview 3.0”), http://www.wipo.int/amc/en/domains/search/overview3.0/.
 WIPO Final Report,¶ 170.
 Id. ¶153.
 Nikon, Inc. v. Technilab. Inc., D2000-1774 (WIPO Feb. 26, 2001).
 Private conversation with panelist Neil A. Brown, QC, referring to Rule 15(a).
 World Wrestling Federation Entertainment, Inc. v. Michael Bosman, D99-0001 (WIPO Jan. 14, 2000).
 Telstra Corporation Limited v. Nuclear Marshmallows, D2000-0003 (WIPO February 18, 2000) (an inference will be drawn from respondent’s choice of name when “it is not possible to conceive of any plausible actual or contemplated active use of the Domain Name by respondent that would not be illegitimate.”)
 Telaxis Communications Corp. v. William E. Minkle, D2000-0005 (WIPO Mar. 5, 2000).
 Allocation Network GmbH v. Steve Gregory, D2000-0016 (WIPO Mar. 24, 2000).
 Smart Design LLC v. Hughes, D2000-0993 (WIPO Oct. 18, 2000).
 EAuto, L.L.C. v. EAuto Parts, D2000-0096 (WIPO Apr. 9, 2000).
 Goldline International, Inc. v. Gold Line, D2000-1151 (WIPO January 4, 2001) (<goldline.com>) citing Smart Design, supra. and Loblaws, Inc. v. Presidentchoice.inc/Presidentchoice.com, AF-0170a to 0170c (eResolution, June 7, 2000).
 City Views Limited v. Moniker Privacy Services / Xander, Jeduyu, ALGEBRALIVE, D2009-0643 (WIPO July 3, 2009) (<mummygold.com>). The new construction was announced in this case, although the Panel denied the complaint. The same Panel later applied his new construction to grant the complaint in Octogen Pharmacal Company, Inc. v. Domains By Proxy, Inc. / Rich Sanders and Octogen e-Solutions, D2009-0786 (WIPO Aug. 19, 2009).
 Time Inc. v. Chip Cooper, D2000-1342 (WIPO Feb. 13, 2001) (<lifemagazine.com>).
 PAA Laboratories GmbH v. Printing Arts America, D2004-0338 (WIPO July 13, 2004).
 Kids & Us English, S.L. v. Target Success.Com, Incorporated, D2016-0356 (WIPO Apr. 8, 2016) (<kidsandus.com>) responding to the argument that the traditional “approach can lead to outcomes which some WIPO panels have considered unjust.” It would only be “unjust” if an owner of a later acquired mark were entitled to a domain name corresponding to its mark, but that is not the law and to assert it would give the mark owner more rights than the law grants.
Published with permission of the New York State Bar Association. Originally published in Vol. 27, No. 2 of Bright Ideas (Fall 2018), a publication of the Intellectual Property Law Section.
Written by Gerald M. Levine, Intellectual Property, Arbitrator/Mediator at Levine Samuel LLP