The new domain name environment and cybersquatting
Download 28.89 Kb.
|
| THE NEW DOMAIN NAME ENVIRONMENT AND CYBERSQUATTING
© 2010 M. Scott Donahey Arbitrator & Mediator 3790 El Camino Real Suite 171 Palo Alto, CA 94306 Tel: 650.823.0338 Fax: 650.941.4262 Email: adr@scottdonahey.com Website: www.scottdonahey.com I. INTRODUCTION The problem of cybersquatting, the confusing, bad faith use of the trademark of a third party by the registrant of a domain name, led in part to the creation of the Internet Corporation of Assigned Names and Numbers (“ICANN”). At the behest of the Clinton administration, prominent members of the Internet community formed a private, not for profit, California corporation to oversee the Domain Name System (“DNS”).1 One specific problem with which the new entity was tasked by the U.S. government, with whom ICANN contracted to provide services in the DNS, was the creation and adoption of a private dispute resolution mechanism to resolve the disputes which were continually arising concerning alleged cybersquatting. Through ICANN’s cooperation with the World Intellectual Property Organization (“WIPO”), the Uniform Domain Name Dispute Resolution Policy (the “UDRP”) was developed and made a part of the contract which every registrar of generic Top Level Domains (“gTLDs”) (such as .com) was required to enter with its domain name registrants. For a period of time, all domain names were required to consist only of ASCII characters.2 In 2003 Internationalized Domain Names (“IDNs”) were introduced into the DNS. Internet Protocol numbers, which constitute the actual address of the various devices which access the Internet, were limited to working with LDH characters, so that there had to be an intermediate step that would translate the non-ASCII characters into LDH (Letters, Digits, Hyphen, a subset of ASCII. Punycode was the translation medium. Still only a very limited number of non-ASCII characters were permitted. Now the experiment in the registration of IDNs will be expanded to include all IDNs that are part of the Unicode system.3 At its inception the DNS was limited to seven TLDs: the three truly generic TLDs (com, .net, and .org) and the four restricted TLDs (.edu, .gov, .mil, and .int.) Gradually it expanded from the original seven to twenty.4 Including the 247 country code Top Level Domains (“ccTLDs”) and the three IDN ccTLDs currently approved, there are some 270 TLDs currently in use. Now the door is open to the general use of IDNs, both in the SLD and as ccTLDs. In addition the unlimited creation and reservation of new gTLDs opens up unlimited possibilities for the gTLD space. This introduces a host of potential problems for trademark owners, not the least of which is the potentially enormous cost in policing the DNS. This paper intends to examine the potential new problems which the new IDNs and the expanded gTLD space pose for trademark owners. II. EXPANDED gTLD SPACE A. Trademarks as gTLDs It is now possible to register a trademark as a gTLD, for example, .alfaromeo. This obviously has some attractiveness to a mark owner, as its online address can now be its trademark, and the mark owner can now organize the online presence of its brands and functions around its mark through the registration of domain names. For example, should Alfa Romeo Automobiles S. p. A. decide to reserve .alfaromeo, it could put information concerning its Spider convertible at B. The Reserved gTLD Is the Equivalent of the Trademark, but Is also Capable of Generic or Descriptive Use Suppose that an unknown organization by the name of Future Apple Farmers of America is the first to reserve .apple as a gTLD? What, if anything, can Apple Records or Apple Computers do? C. Trademark Plus Common Word or Phrase as gTLD Someone may also choose to reserve a trademark plus a common word or phrase as a gTLD. For example, someone may reserve .alfaromeosportscars. If that someone is the trademark owner, the trademark owner may be very well pleased. However, the party reserving the gTLD may be a third party or a competitor seeking to profit from the good will associated with the mark through the registration of an SLD in the new gTLD, for example D. gTLD as Homonym of Trademark It is possible to reserve a gTLD that is a homonym of a famous mark. Two examples of such a potential reservation are .alpharomeo and .alfaromayo. The UDRP was designed to deal with registrations in the SLD which created confusion, and did not contemplate the possibility that a trademark or variations thereon might be reserved as a gTLD.
Typosquatting is already a serious problem in SLD domain name registrations. An example is F. gTLD as Generic Term Even the reservation of generic terms as gTLDs vastly increases the potential of cybersquatting: .auto, .vehicle, .sportscar, and .voiture. are all natural gTLDs in which to find the registration of an SLD related to the automotive industry, such as “alfaromeo.” The registrant of the domain name III. THE REGISTRATION OF INTERNATIONALIZED DOMAIN NAMES (IDNs) A. Existing Use of IDNs 1. Prior to July 2003 As previously indicated in the introduction, prior to 2003, IDNs were not possible of registration in the domain name system and the entry of non-ASCII characters was simply ignored. However, “internationalized domain names” (“idns”) were in fact in use, because it was possible to register non-English words in SLDs, so long as only LDH characters were used. To the extent that non-ASCII diacritical marks were entered in a browser by a user, they were simply ignored. Cybersquatters recognized an opportunity to profit from famous marks simply by literally translating the mark into another language where the foreign translation of the mark had not been registered. This can be seen in a short discussion of the UDRP decision, Société pour l”Oeuvre et la Mémoire d’Antoine de Saint Exupéry – D’Agay v. Perlegos Properties, WIPO Case No. D2005-1085, ( In The Little Prince case, the domain name registrant had registered the domain name 2. July 2003 to ICANN’s Implementation of IDNs In June 2003, ICANN issued guidelines for use of IDNs in SLDs. Registration of IDNs in .jp began in July 2003. In March 2004, several registrars of TLDs began to register IDNs in SLDs. Punycode was employed to enable the registration of IDNs in the SLD. Punycode is a particular subset of LDH which provides for the translation of an IDN into an LDH equivalent. When confronted with a domain name containing at least one non-ASCII character in an SLD, an algorithm called “To ASCII” will apply various rules to translate the IDN into ASCII characters. The four character prefix string, referred to as Ace Prefix, is used to distinguish punycode encoded strings from ordinary ASCII strings. Since it was impossible to register IDNs in a domain name prior to 2003, non-English words using diacritical marks, such as an accent mark, had to be registered without the mark. Thus, although computers and computer programs existed that permitted the typing and use of non-English words with diacritical marks in word processing programs (and therefore capable of entry into browsers), those same diacritical marks could not be reflected in a domain name. Therefore, domain names registered prior to the use of punycode would simply ignore the diacritical mark when entered from a computer capable of keying in such a mark in an Internet browser. However, after the introduction of punycode, the DNS would recognize the punycode translation of the IDN and take the user to the web site to which the punycode translated domain name resolved. This opened the door for cybersquatters to prey on trademarks that consisted of non-English words which included at least one non-ASCII diacritical mark. This can be illustrated by a brief discussion of the UDRP case Monografias.com S.A., Fernando Julián Negro v. Nick Lozikov / Moniker Privacy Services, WIPO Case No. D2008-0140 ( In Monografias the complainant registered the domain name Respondent registered the punycode translation of the domain name 3. The new IDNs and Unicode In the mid 1980s, work began on the development of a uniform, universal code to represent the varying language, numerical, and special characters in use in languages throughout the world (and more recently those of historical languages no longer in use). Dubbed “Unicode,” the results of the efforts of largely American high tech companies resulted in the first release (Version 1.0.0) in October 1991. The most recent release occurred on October 1, 2009, and was Unicode Version 5.2.9. The list of linguistic characters included in Unicode is approaching 100,000 unique characters encoded. a. Graphic confusion in ASCII Although graphic confusion rarely exists in ASCII, and even more rarely in LDH, with certain homographs and in certain fonts confusion can arise. Upper case “O” and “0” are sometimes confused, especially in certain fonts. Likewise, lower case “l” and the numeral “1” can be confused for each other. In certain fonts the combination of the lower case characters “r” and “n” (“rn”) closely mimics the letter “m.” As discussed, infra, unscrupulous users can take advantage of these confusing similarities. b. Graphic confusion in Unicode The addition of thousands of non ASCII characters simply increases the possibility of confusion among homographs. For example the Cyrillic character “а” is virtually identical to the Latin character “a.” The Cyrillic alphabet contains 10 other lower case characters that are identical or very similar to Latin characters. Moreover, Cyrillic numerals з, ч, and б closely resemble Arabic numerals 3, 4, and 6. The Korean language borrowed Chinese ideographs and made them Korean. Korean ideographs are called hanja. But the Koreans also developed a character set that represents sounds. The Korean character set, hangeul, is based on the pronunciation of the Korean language, hanmal. Each character represents a syllable and corresponds to a sound, rather than to an idea. Each character conveys no meaning, but rather only the sound of the particular syllable. Thus it is possible to arrange hangeul so as to mimic the sounds of non-Korean words. In this regard, the Korean language is similar to the Japanese language in which kanji and kokuji are graphic and ideographic forms of the language, while hiragana and katakana are scripts based on pronunciation, similar to Korean hangeul. Hiragana and katakana are used to integrate foreign words or phrases into the Japanese language. Arabic and Hebrew writing systems are bidirectional. Text runs from right-to-left and numbers or embedded Latin characters are read from left –to-right. These writing systems are known as “bidi.” This is potentially another source of possible confusion. c. The problem of homograph spoofing attacks Through the use of similar looking, but different, characters, someone could register a domain name that mimics an already registered domain name which is used to resolve to a particular web site. However, when one clicks on the similar looking name, one is taken to a different web site. The creation of the illusion that one is going to site X, when in fact the similar character string is taking one to site Y by the use of similar appearing characters is known as homographic spoofing. That second site may also mimic the look and feel of the prior web site, and may be used to induce the user to disclose personal information (passwords, bank accounts, identification information, etc.) that may be used in various illegal ways. This is generally known as phishing. The use of Unicode greatly increases the possibility of homograph spoofing attacks and of phishing. IV. CONCLUSION The current system for dealing with the problem of cybersquatting is inadequate to deal with the new DNS environment, an environment in which TLDs can be virtually anything one is willing to pay to reserve and which can be composed of IDNs, some of which may closely resemble existing TLDs . (For example the reservation of a gTLD consisting of the Cyrillic “с,” the Greek “ο,” and the Latin “rn,” all in the Tahoma font, would look like this: “.сοrn.”) The costs involved in reserving and or challenging all the possible variations of a trademark that might be used in conjunction with a new gTLD or which might be reserved as a new gTLD are staggering. The use of IDNs greatly increases the potential for homograph spoofing attacks and phishing. There not only needs to be new procedures to resolve multiple disputes at one time, rather than piecemeal, but also procedures which inhibit registrars from continuing to register obviously infringing names. But there also needs to be new technical solutions to the problem of the similarities in different character strings and the ability to distinguish the characters in one linguistic system from those of another linguistic system. Finally, some thought should be given to the possibility of standardizing one font for Internet use. 1 The use of trademarks in the DNS differs markedly from their use in national (and international) commerce. First, in the geopolitical world, trademarks are the creation of the laws of a sovereign and as such they are generally limited to recognition and enforcement to the territory of the sovereign creator. A domain name is present everywhere at once on the Internet. In the civil and common law regimes, trademarks are limited both by the geographical areas in which they are in use and the products or services with which they are used. Thus, United Air Lines and United Van Lines can peacefully co-exist in the trademark world, just as McDonald’s Lorry Services in Glasgow can peacefully co-exist with the McDonald’s restaurant empire. Domain names are issued by private registrars and their availability is necessarily limited: there can only be one owner of united.com or mcdonalds.net. 2 The American Standard Code of Information Interchange consists of one hundred twenty-eight alpha, numeric, and special characters, including ten Arabic numerals 0-9, fifty two Roman alphabetical upper and lower case characters, and the balance of special characters. Domain names have been further limited to Letter Digit Hyphen characters (“LDH”), consisting of the twenty-six lower case Latin alphabet, the ten Arabic numeral characters, and the hyphen, “-“. It should also be noted that the dot (“.”) is used as a divider between discrete levels of TLDs or to indicate the beginning and/or end of semantic units. 3 Unicode, which will be discussed in more detail, infra, is a project of the Unicode Consortium, a non-profit organization largely consisting of corporations and organization in the computer and information processing sectors. Its goal is to establish a unique number for every character, no matter what the platform or language. 4 The unrestricted new gTLDs include .biz, .info, name, and .pro, with the balance either restricted or sponsored (.aero, .coop, .museum, .asia, .cat, .jobs, .mobi, .tel, and .travel. Directory: pdfs pdfs -> The petition and case system pdfs -> Sal men’s 12 Step Recovery Meeting for Sexual Addiction Introduction pdfs -> What is ascaa? pdfs -> Tel./Fax No. 2287 3696 pdfs -> 1–xx. The Redevelopment of a Warm Core Structure in Erin: a case of Inland Tropical Storm Formation pdfs -> Policies and Procedures 40 – Eligibility pdfs -> Policies and Procedures 40 – Eligibility pdfs -> Rev. 10 Sep 2008 Ozark Hispanic Studies League pdfs -> Handbook of exercises for transportation sector personnel Download 28.89 Kb. Share with your friends:
|