Comment on Closed Generic TLDs (LIPTON)

["Lipton, Jacqueline D" <jdlipton@xxxxxxxxxxxxxx>]

March 6, 2013

“Closed Generic” TLD Applications 

I write in response to ICANN’s call for public comments on the issue of “closed 
generic” TLD Applications dated 5 February 2013.

I am a professor of law whose expertise resides in the field of law and digital 
technology, with particular focus on trademarks, free speech, and Internet 
governance.  I am the author of “Internet Domain Names, Trademarks, and Free 
Speech” (Edward Elgar, 2010) as well as more than forty law review articles 
dealing with Internet domain name disputes and Internet governance more 
generally.  I also co-author the leading American Internet law casebook, 
“Cyberspace Law:  Cases and Materials” (2 ed, 2005, and 3 ed, 2010) (with R. 
Ku).  I hold Ph.D.s from both Cambridge University and Griffith University, 
both in the fields of regulating digital technology.  My teaching is focused in 
the areas of International Intellectual Property, International Business 
Transactions, Trademark and Unfair Competition, and Internet Law.

ICANN has asked for comments on the issues of developing workable criteria for:

•       classifying certain applications as “closed generic” TLDs, i.e., how to 
determine whether a string is generic, and
•       determining the circumstances under which a particular gTLD operator 
should be permitted to adopt “open” or “closed” registration policies.

In my view, ICANN should not adopt any new policies for what might be termed 
“generic” TLDs.  Adopting a new policy for generic TLDs at this point would be 
inconsistent with the way the system was conceived, developed, and publicized 
by ICANN.  Much care and thought were put into creating the system.  Potential 
issues concerning generic gTLDs were considered by ICANN and others during this 
time.  Applicants for new gTLDs have relied on the criterion as originally 
stated by ICANN.  Changing course now would be inequitable and cause more cost 
and inconvenience both for ICANN and for those who applied for new gTLDs based 
on the original application criteria.

My detailed reasons in support of this conclusion are set out below.

1.      As ICANN has acknowledged in its call for public comment, no specific 
rules have been implemented for assessing applications for generic as opposed 
to commercial brand-related TLDs.  A number of applicants have relied on the 
initial representations by ICANN in formulating business plans to apply for 
what might be termed “generic” TLDs.  Many of these business plans will likely 
further ICANN’s stated objective of “encouraging innovative business models”.  
These applicants have invested significant resources in making their 
applications, and a number of them have developed innovative business plans 
that require them to run “closed” registries if their applications are 
successful.  Such plans are consistent with the way ICANN established the 
process.  Imposing a requirement that such applicants now incur the costs and 
resources involved in maintaining an “open” registry would create an 
inequitable burden and would hinder them in achieving their business goals.  
Indeed, such a change of course in policy by ICANN may, in fact, expose it to 
the possibility of litigation from those who relied on its initial 
representations which were silent on the issue of generic versus commercial 
brand-name TLDs.

2.      While much of the concern around closed generic TLDs has centered on 
currently applied-for terms which are fairly obviously “generic” (such as 
.free, .shop, and .search), if ICANN were to adopt a new set of guidelines for 
generic TLDs, it would be extremely difficult, if not impossible, to devise a 
workable definition for “generic” in this context.  Within an international 
system like the domain name system, it is likely to be the case that terms 
considered “generic” in one culture or country may not be generic in another.  
Additionally terms considered “generic” in one context may be brand names in 
another.  Geographical names are a good example.  Currently the trademark 
holder for “Patagonia” sporting goods has applied for “.patagonia”.  This term 
could be considered generic or at least geographic.  Many extremely powerful 
American trademarks are also generic words in the English language: for 
example:  “gateway”, “dell”, “playboy” and “playmate”.  

ICANN would also potentially face difficulties of definition with terms that 
are commercial brand names in one context, but are generic symbols, such as 
Greek letters, in another.  “Delta” is a good example of a term that is 
trademarked by multiple commercial entities as well as connoting a linguistic 
and mathematical symbol.  There is no effective way for ICANN to define 
“generic” in a way that would make sense for the international community.  
ICANN would face difficulties, and potentially also litigation from aggrieved 
trademark holders, if it attempted to define “generic” in this context and to 
impose a rule that any character string regarded as generic would have to be 
operated as a closed registry.  

3.      ICANN has historically taken a “hands off” approach to policy-making in 
the domain space, outside of protecting trademark holders against bad faith 
conduct, typically by cybersquatters.  This approach is consistent with ICANN’s 
by-laws which contemplate that it will only make policy determinations with 
respect to the technical aspects of the system.   Straying into more general 
policy-based rulemaking for the use of specific words and phrases would be 
inconsistent with ICANN’s by-laws as well as inconsistent with previous 
approaches taken to regulating domain spaces.

4.      If ICANN chooses not to formulate specific rules for new generic TLDs, 
the task of resolving concerns about unfair and inappropriate conduct within 
the new domain spaces will fall to domestic legal systems.  This is consistent 
with ways in which existing domain name disputes have been handled, outside the 
Uniform Domain Name Dispute Resolution Procedure (UDRP) which is itself subject 
to national court proceedings.  

Institutions that have applied for generic TLDs under the new process will not 
be immune from litigation in domestic courts related to issues like antitrust, 
trademark infringement, consumer fraud etc.  These issues are more 
appropriately dealt with by domestic courts and legislatures with the 
wherewithal and constitutional authority to deliberate on them than on a body 
whose purpose is to administer the technical aspects of the system.

5.      Arguments have been raised that if ICANN fails to adopt special rules 
for new generic TLDs, online innovation will be chilled.  There is simply 
insufficient evidence to make this case convincingly.  Innovation can take many 
forms. There is no empirical way of proving at this point in time whether 
innovation would be encouraged by allowing the operators of new generic TLDs to 
operate them as closed registries or by forcing them to incur the time, cost, 
and associated responsibilities of operating them as open registries.  If 
applicants are ultimately permitted to maintain a new gTLD as a closed 
registry, they will be free to innovate in ways that may not be possible if 
they were obliged to maintain an open registry.  Further, those that fail to 
obtain access to a given new generic TLD may find new ways to innovate in other 
domain spaces, thus further promoting ICANN’s objectives of “fostering choice, 
innovation, and competition”.  

6.      As a corollary to the previous point, there is no scarcity of potential 
gTLDs, and allowing applicants to pursue their legitimate business plans as 
stated will not necessarily hinder innovation online.  Others will remain free 
to pursue new business plans in other domain spaces.  This is consistent with 
the situation under the existing domain name system where the fact that one 
entity holds a second-level domain that another entity might like to use does 
not prevent that second entity from innovating in another domain space, or from 
negotiating with the original domain name holder for the transfer or use of the 
desired domain space.  The market has been left open to organize itself in 
existing domain spaces.  This policy should be carried over to new gTLDs.

7.      ICANN is not obliged to grant a generic TLD to any one particular 
applicant.  If, for example, there is a debate about which entity should be 
granted a specific new gTLD such as “.shop”, there are procedures in place to 
ascertain priority, including ultimately an auction between applicants 
contesting the right to the same character string.  Once the established 
procedures have been carried out, ICANN’s role ceases, unless the applicant 
runs the registry inconsistently with the business plan submitted along with 
the application.  This should be the end of ICANN’s role in the matter as 
stated in the current Applicant Guidebook.

8.      Some commentators on new generic TLD applications have made specious 
analogies with the problem of cybersquatting when the domain name system was 
first introduced.  Cybersquatting involves the potential waste of a useful 
resource as the result of bad faith conduct by a registrant the purpose of 
which is to extort money from a legitimate trademark holder.  Unlike 
cybersquatting, the application by an entity to apply to run a closed generic 
TLD registry involves no such motivation or negative effect on online 
innovation.  Applicants for these gTLDs have worked to develop innovative and 
legitimate business models, not to hold otherwise useful domain addresses for 
ransom to the highest bidder.  

9.      Additionally, and as a corollary to the previous point, the approach 
ICANN took to the problem of cybersquatting was to develop the UDRP.  ICANN did 
not “regulate” the initial grant of any domain names in existing domain spaces 
to deal with the perceived threat of cybersquatting.  Rather it imposed a 
private contractual arrangement on registration of a new domain name within 
certain gTLDs to ensure that trademark holders had a cost-effective avenue for 
dealing with bad faith cybersquatters.  By considering rule-making that would 
impact who could register a new gTLD ex ante under the new gTLD process, ICANN 
would be acting inconsistently with its past practices and procedures.  

The UDRP allows legitimately aggrieved parties an effective avenue to protect 
their existing interests in the domain space.  In comparison, a rule that would 
require applicants to maintain open, rather than closed, registries for generic 
character strings would assume the existence of aggrieved parties before the 
fact and would impose a burden on applicants inconsistent with any prior 
development in the domain space.
In sum, I would urge ICANN to maintain consistency with its stated procedures 
for applications for new generic TLDs and not to impose new rules respecting 
“closed” versus “open” new gTLD registries.  Applicants have relied on ICANN’s 
initial representations about how the system will work, often at significant 
cost.  There was adequate time during the development of the system to consider 
the issue of generic TLDs and it is not appropriate now that applications have 
already been solicited to change the rules.  There is no evidence that allowing 
closed registries under generic character strings would hinder, rather than 
facilitate, online innovation.  Further, entities that are successful in 
obtaining generic TLDs will be subject to local laws including antitrust, 
fraud, trademark etc.

Thank you for the opportunity to comment on this issue.  If you have any 
queries in relation to the above, my contact details are below.


Sincerely,
Jacqueline D. Lipton, Ph.D.
Baker Botts Professor of Law
Co-Director, Institute for Intellectual Property and Information Law
University of Houston Law Center





Jacqueline D. Lipton, Ph.D.
Baker Botts Professor of Law
Co-Director, Institute for Intellectual Property and Information Law
University of Houston Law Center
100 Law Center
Houston TX  77204-6060

Tel:  713 743 1543