From: Jonathan Moskin <moskinj@PENNIE.COM>
Date: 4/16/98 4:46pm
Subject: Comments Re NTIA Green Paper
I am a partner in the intellectual property law firm, Pennie &
Edmonds LLP. Attached (in WordPerfect 5.1) are comments regarding
the NTIA Proposal To Improve Technical Management of Internet
Names and Addresses. The comments reflect my own views and are
not necessarily those of the firm or any of its members.
I am grateful for the opportunity to submit these comments on
issues of potentially great legal and practical significance.
Jonathan E. Moskin
Comments Regarding NTIA Proposal To
Improve Technical Management of Internet Names and Addresses
It is a bedrock principle of trademark law that rights are territorial. Famous marks aside, use or registration in one jurisdiction do not confer rights (or, generally, liabilities) elsewhere.(2) The Internet by contrast, is without territorial boundaries. Trademark law also permits identical marks to be used simultaneously in different territories, or even by two or more non-competing businesses within any given country, provided there is no likelihood of confusion.(3)
"Apple," thus, can serve simultaneously as a trademark for computers, for a bank, or for sound recordings. The Internet by contrast, has not been nearly so flexible, since within any given top level domain (for instance, the top level domain ".com") only one party may own a given domain name.
Through various proposals, the internet community has sought to resolve the conflicts among parties competing for rights in a single domain name (or one or more similar names), and open up to competition the business of registering domain names -- a business that to date has more or less been monopolized by Network Solutions, Inc. under contract with the National Science Foundation. There have been two main proposals for reforms have been the "Green Paper" of the U.S. Department of Commerce's National Telecommunications & Information Administration (NTIA), and the "gTLD Memorandum of Understanding" of the Internet Counsel of Registrars (CORE). Although both proposals are meant to remedy trademark conflicts, neither have been developed with basic principles of trademark law fully in mind.
Although the NTIA and CORE proposals differ, both involve multiplying the number of generic top level domains (gTLD's) and increasing the number of domain name registrars, thus ending the exclusive role served to date by Network Solutions in registering domain names in the existing gTLD's, .com, .org and .net. Under the NTIA Green Paper, there would be five new (as-yet-undefined) gTLDs; under the CORE proposal, there would be seven: .firm (for businesses or firms); .shop (for businesses offering goods to purchase); .web (for activities related to the World Wide Web); .arts (for cultural and entertainment activities); .rec (for recreation and entertainment activities); .info for information services); and .nom (for personal uses). Like the top level domain ".com", the proposed new gTLD's will cut across international boundaries.
Although the NTIA proposal is drafted very broadly, and hence could be
implemented in varying ways, it appears (from the perspective of the trademark owner or domain
name registrant) to share with the CORE proposal four main difficulties: a likely increase in
international conflicts over individual domain names; a likely increase in international disputes
over conflicts of laws and extraterritorial assertions of jurisdiction; an increased need to police
new top level domains, and insufficient attention to the public, quasi-governmental role of domain
name registry services.
Internationalizing domain name registries means that, within any given top level domain, the entire world will be competing for rights in any given name. Even within any one country, conflicts have arisen because two or more noncompeting businesses that can co-exist lawfully using the same trademark (e.g., PRINCE pasta and PRINCE sporting goods) may not simultaneously use the same domain name. However, the likelihood of finding two or more entities legitimately using the same or similar names can only increase when the relevant field is expanded from one country to the entire globe. Adding five or seven new gTLD's can hardly offset the likely geometric increase of conflicts from making all gTLD's international in scope. Moreover, many businesses will likely seek to establish rights in more than one (and may be all) of the new gTLD's. By insufficiently acknowledging trademark principles of territoriality, the Green Paper and the CORE proposals thus seem likely to encourage new and potentially needless international conflicts among legitimate businesses.
Without sacrificing global interoperability of the Internet, the domain name registry system need not itself be completely borderless. There are already 241 separate country-specific or "national" top level domains, such as "us." for the United States, ".de" for Germany, ".ne" for the Netherlands, and so on. Although the ".us" TLD has not been widely used, and although it may be very difficult to re-orient American or other businesses already engaged in commerce in the gTLD ".com" to the national TLD ".us", such a geographically-oriented system has obvious advantages. The most obvious such advantage is that a system of national registries would be in harmony with trademark principles of territoriality. Under a system of national top level domains, disputes over domain names thus could be decided under the domestic laws of the chosen country. And where litigation is necessary to resolve rights in domain names, issues of extraterritorial jurisdiction and foreign service of process are less likely to arise. Such issues need never arise if registration in a national TLD also required appointment of an agent for service and consent to jurisdiction. The NTIA proposal notes that increased use of the ".us" domain could help resolve existing domain name conflicts, and the original proposal of the Internet Ad Hoc Committee (under whose aegis the CORE proposal was developed) likewise favored increased use of the .us domain as a means of reducing conflicts in the .com domain.(4)
Although industry-oriented gTLD's (e.g. .firm, .store, .arts) that the CORE proposal and the NTIA Green Paper favor may help ease such conflicts among trademark owners, the addition of new domains could not possibly erase all such problems -- nationally or globally. Very simply, the internet will never be free of the types of trademark conflicts that have always existed in ordinary commerce. Instead, within any given national or generic TLD, a more subtle "directory" system, such as the system implemented at the "lloyds.io" website, may offer a technically simple and legally elegant solution. Under a directory system, multiple legitimate users of the same name (e.g. APPLE records, APPLE computers or APPLE bank) all can share a single directory page on the Internet, from which Internet users would be directed to the appropriate site if they prefer bank records to Beatles' records.(5) Such an orderly presentation of names, with explanatory notes as to the identities of the respective domain name owners, would unquestionably be more helpful in preventing confusion than would happenstance search results obtained by Internet users using conventional search engines, such as Yahoo or AltaVista, scanning multiple industry-oriented gTLD's. Trademark owners may chafe at the prospect of sharing a single page with unrelated third parties using the same name. However, if such an owner has a basis for claiming truly exclusive rights in its name, it may take appropriate legal action against the perceived infringer. If the trademark owner does not possess exclusive rights in its name, it will most likely have little basis to complain. Within a national directory system, the owner would have at its disposal the full panoply of rights to preserve any claim it may have to exclusivity -- no less, but no more.
The value of industry-oriented domains (such as .store or .arts) might still be accommodated within a geographically-oriented system by using such registries as second level domains (e.g., ".us.store"). Moreover, since the distinction of principal relevance in trademark law is between commercial and non-commercial uses, many benefits are apparent from a top level -- or, within a national system, second level -- domain such as the ".nom" suggested by CORE for personal (non-commercial) use. The top level domains ".edu" and ".gov" are also sufficiently non-commercial to justify separate treatment in domain name registration and dispute resolution procedures.(6) Specially protected First Amendment activities, such as news gathering and publication, might also merit separate treatment in a .news domain.
Although much of the business transacted on the internet is national or
international in its reach, much is also purely local. Individuals establish websites to interact with
friends; local businesses advertise their products or services locally, and local communities create
forums for public comment on local issues. Within any one country, regional second or third level
domains (e.g., .us.ny) could be used for internet sites of distinctly local appeal, such as local
school board issues or local restaurants or jazz clubs. By designating domain names and websites
as having a principally local appeal, purely local users might also be less subject to (although not
entirely free from) suit in remote jurisdictions. Considerable ink has already been spilled
determining when operation of a website may subject a business to suit in remote jurisdictions.(7)
Given the extent to which business and internet users have come to populate the
top level domain .com, it may be difficult to put the genie back in the bottle and shift from an
international registry system to a country-specific system -- or to the proposed new gTLD's
contemplated by the Green Paper or CORE proposal. Indeed, just as the .us top level domain has
been under-utilized to date, it may be that the proposed new gTLD's will be largely ignored in
favor of the existing gTLD .com. Any such change in orientation is a bit like trying to board a
moving bus.(8) Nor can any such reorientation sacrifice the international interoperability of the
internet. However, website owners might be encouraged to transition to national registries if it
were to become a condition to registering in the top level domain ".com" (or other gTLD's) that
the registrant consent to jurisdiction everywhere in the world. A further incentive would be to
spare domain name registrants from such a potentially onerous result when registering in a
national TLD only.
The Green Paper endorses the value of domain name dispute resolution policies, and contemplates that each new gTLD registry have its own such policy. No specific procedures have yet been proposed. The CORE proposal, by contrast, contemplates having all registries follow one dispute resolution policy. That policy has been developed in some detail.(9) It includes on-line submissions under procedural rules established by WIPO (the World Intellectual Property Organization), based in Geneva, Switzerland, and a set of substantive principles developed by CORE to balance the interests of trademark owners with the need for fair administration of domain name registries.
CORE's proposed rules are obviously the product of careful thought. Nonetheless, the substantive and procedural rules proposed by CORE do not reflect the laws of any specific country. Trademark owners in any given country thus may be concerned that the CORE dispute resolution system will yield results contrary to those under their own national law. Moreover, although CORE contemplates a right to appeal (within the CORE system), the CORE mediators and arbitrators are not accountable under the laws of any specific jurisdiction. No matter how well-intentioned the CORE proposal, trademark owners thus cannot be certain how the adjudicative power given to such internationally-at-large arbitrators will be wielded -- or what recourse they may have for abuses of process.
Parties who do not consent to adjudication of valued domain name rights by potentially unaccountable tribunals applying an unknown and uncertain body of substantive and procedural law, may choose instead to litigate in familiar forums. By internationalizing the dispute resolution system, the new proposals thus may engender new international litigation, with new disputes concerning personal jurisdiction over foreign domain name registrants, as well as conflicts of laws questions over the substantive principles and precedents to apply in deciding such cases. Within the United States, such jurisdictional disputes have been extensive(10). Some of the cases already have had international ramifications.(11)
Although the CORE proposal acknowledges the possibility of litigation outside its preferred dispute resolution system, the proposal does not address whether concerns over the accountability of the CORE adjudication system or the substantive law applied may in fact encourage international litigation. Certainly CORE could not prohibit suits in U.S. federal court by aggrieved trademark owners against foreign domain name registrants for direct infringement -- or, quite possibly against foreign-based registrars or registries for contributory infringement. Under U.S. law and the extra-territorial reach of the Lanham Act, such suits may be maintained against foreign parties upon a showing of substantial impact on U.S. commerce.(12)
Such suits may likewise be maintained abroad against U.S. domain name owners or U.S.-based domain name registrars.
Under a geographically-based domain name system, there would be less need for
international dispute resolution. Disputes could be resolved under the domestic laws of each
respective country. There would be less need for extraterritorial service of process - and no need
for extraterritorial service of process if registration within any national top level domain required
appointment of a domestic representative or consent to jurisdiction and service of process.
As trademark owners increasingly have come to protect their most valued trademarks by registering them in the top level domain .com, the number of domain name disputes in the .com gTLD seems likely, ultimately, to decline. However, the creation of five or seven (or maybe more) new gTLD's, creates whole new frontiers for piracy. Indeed, any system that multiplies the number of top level domains (as both the CORE and NTIA proposals do) multiplies as well the burden on trademark owners to police and enforce their rights -- unless perhaps the number were to be multiplied to such an extent as to diffuse any special value of establishing rights in a given gTLD. Under the CORE and NTIA proposals, each new gTLD will be a potential new venue for domain name piracy, and each will require policing.
Creation of new gTLD's also will not help prevent likelihood of confusion. It
may even facilitate confusion. Because of the complete interoperabilty of the Internet, the
existence of identical names in any of the proposed gTLD's will be apparent when the exact name
(or a similar name) is run through any of the available Internet search engines. Cutting across
existing and proposed new gTLD's, the search results obtained from Yahoo, MetaCrawler or
AltaVista, or any other search engine, will immediately expose internet users to the fact that there
may be not only an "apple.com", but also an "apple.firm", an "apple.shop" or an "apple.arts".
Increased reliance on national TLD's will not necessarily help resolve this problem - particularly if
industry-specific or generic second level domains were incorporated into a national system. The
only lesson here is to proceed cautiously in establishing any new domains -- or move to a far more
open system under which none of the gTLD's have any special value. The limited number of new
registries may be a boon to the businesses that may operate them at a profit yet a bane to
trademark owners who must police them at a loss.
The internet community is rightly covetous of the monopoly power and financial status enjoyed by Network Solutions as the exclusive registry of gTLD's. However, the great eagerness to create competition in the registry of domain names has not been accompanied by any clear acknowledgment of the essentially public, quasi-governmental, function of domain name registration. The CORE proposal notes in passing that the Internet must be administered as a public trust, but specifies no duties that follow from this trusteeship. The NTIA Green Paper does not address the duty, if any, registrars or registries may owe the public generally, and specifically disavows any public duty domain name registrars owe to trademark owners. Indeed, it suggests without analysis that there is no such duty.(13)
Internet service providers have not been immune from liability for contributory copyright infringement(14) or defamation.(15) The Lanham Act likewise recognizes a right of action for contributory infringement(16). There is thus no reason that domain name registrars, who in their private capacity enjoy an income stream from registering domain names, should be immune from suit for public injury caused by facilitating acts of trademark infringement or the failure to protect adequately the legal and proprietary interests of domain name owners.(17) Network Solutions, which has been named in many infringement suits, tacitly acknowledges this duty by requiring new domain name applicants to indemnify it from liability.(18) For the domain name registry system to be improved, not simply transitioned from Network Solutions to new registrars or registries, it must become fully accountable, and the public duties of private registrars and registries must be recognized explicitly.
It is, of course, politically incorrect to suggest direct government regulation of the last great frontier, cyberspace. However, the Internet may, in this respect, already be a victim of its own success. It may have been one thing to shun regulation when the Internet was simply (or principally) a means of communication for academicians and forward-thinking computer lovers. It is another matter entirely to continue to maintain a hands-off attitude when the Internet has been transformed into a multi-billion dollar commercial venue. Certainly the author is aware of no careful analysis whether the government could administer a domain name registry system more cost-effectively and fairly than private firms, and with greater public accountability. Nor has there been any public advocate for the possible advantages of exploiting existing governmental expertise in maintaining -- and resolving disputes over -- rights in names. Such expertise resides in the United States Patent and Trademark Office and similar government trademark offices worldwide.
The Supreme Court has quite rightly recognized that the Internet is a "unique and wholly new medium of worldwide communication" with virtually limitless avenues for expression.(19) Unlike traditional broadcasting, which requires allocating limited radio or television frequencies, the internet to date has not required or been subject to extensive regulation. However, if the need for government regulation and international conflict are to be minimized, and if the public is to reap the full benefit of a transition away from Network Solutions' current monopoly over the existing gTLD registries, private firms that seek to assume the public function of maintaining and adjudicating rights in domain names must be fully accountable to the national and worldwide communities they would serve. Because of the difficulty of ensuring accountability internationally, the national model must be preferred.
1. Jonathan E. Moskin is a partner at Pennie & Edmonds LLP. He can be contacted at firstname.lastname@example.org.
2. Buti v. Impressa Perosa S.r.l., 1998 U.S. App. LEXIS 2875 (2d Cir. Feb. 24, 1998).
3. Interstellar Starship Services, Ltd. v. Epix, Inc., 983 F.Supp. 1331 (D. Or. 1997).
4. See www.gtld.mou.org/draft-iahc-gtldspec-OO.
5. See M. Sartori, A Proposal For The Registration of Domain Names, 87 Trademark Rptr. 638 (1997).
6. There is also some appeal (that any parent can perceive) to creating a separate gTLD for pornography -- a seemingly simple way of segregating the seamier side of the internet. However, First Amendment considerations likely render unworkable the sort of line-drawing necessary to encapsulate pornography. Reno v. ACLU, 117 S.Ct. 2329 (1997). Many minors too might welcome such a simple device for locating exactly what their parents disapprove of.
7. Compare Bensusan Restaurant Corp. v. King, 126 F.3d 25 (2d Cir. 1997), and Weber v. Jolly Hotels, 977 F. Supp. 327 (D.N.J. 1997); with Panavision Int'l v. Toeppen, 938 F. Supp. 616, 618-22 (C.D. Cal. 1996), and Maritz, Inc. v. Cybergold, Inc., 947 F. Supp. 1328 (E.D. Mo. 1996). See also Hasbro, Inc. v. Clue Computing Inc., 45 U.S.P.Q.2d 1170 (D. Mass. 1997).
8. See Bensusan Restaurant Corp. v. King, supra note 6, at 126 F.3d at 27.
9. See gtld-mou.org/docs/tracps.
10. See, supra note 6.
11. Playboy Enterprises, Inc. v. Chuckleberry Pub. Co., 939 F.Supp. 1032 (S.D.N.Y. 1996) (New York Court with prior jurisdiction over Italian publisher enjoined the foreign company from making available to internet users in the United States its Italian on-line publication bearing an infringing trademark. See also Weber v. Jolly Hotels, supra note 6 (claim in New Jersey Federal District Court against foreign hotel chain based on alleged U.S. presence via the hotel's website).
12. Steele v. Bulova Watch Co., 334 U.S. 280 (1952) ; Vanity Fair Mills, Inc. v. T. Eaton Co., 234 F.2d 639 (2d Cir.), cert. denied, 351 U.S. 871 (1950). See generally, J. Dabney, On the Territorial Reach of the Lanham Act, 83 Trademark Rptr. 465 (1993).
13. The Green Paper thus embraces the dubious proposition that "U.S. trademark law imposes no general duty on a registrar to investigate the propriety of any given registration." Cf. infra note 15 and accompanying text.
14. See generally Religious Technology Center v. Netcom On-line, 907 F.Supp. 1361 (N.D. Cal. 1995), and cases cited therein.
15. See Stratton Oakmont, Inc. v. Prodigy Services Co., 23 Media L. Rep. 1794, 1995 N.Y. Misc. LEXIS 229 (Sup. Ct. Nassau County May 24, 1995) (Prodigy liable as "publisher" of on-line bulletin board).
16. Inwood Laboratories, Inc. v. Ives Laboratories, Inc., 456 U.S. 844 (1982); Fonovisa, Inc. v. Cherry Auction, Inc., 76 F.3d 259 (9th Cir. 1996).
17. One court has said that Network Solutions was not liable for direct or contributory infringement for registering a domain name identical to the plaintiff's mark SKUNK WORKS. In so holding, the court noted the indistinctiveness of the mark and the resulting difficulty of requiring Network Solutions to determine rights in the mark. Lockheed Martin Corp. v. Network Solutions, Inc., 985 F.Supp. 949, 963-4 (C.D. Cal. 1997). Although the court declined to hold that Network Solutions' receipt of registration fees satisfied the commercial use requirement of the Lanham Act, it did not consider whether receipt of such fees entailed some affirmative duty to take reasonable steps to avoid infringement, such as investigating the federal trademark registry, requiring appropriate assurances of non-infringement from domain name applicants or requiring notice to registered trademark owners. Nor did the court assess whether Network Solutions' continued maintenance of the domain name registration after obtaining actual knowledge of the plaintiff's mark was contributory infringement.
18. Domain Name Dispute Policy (Rev. 03), ¶ 6, Feb. 25, 1995; see www.netsol.com/rs/dispute-policy.
19. Reno v. A.C.L.U., 117 S.Ct. 2329, 1997 US Lexis 4037 at *67 (June 26, 1997).