I am Senior Counsel for marketing
legal issues, as well as Director of Trademarks and Brands, for Intel Corporation.
I am commenting on several of the issues raised on the question of adding new gTLDs.
First, it is not clear that there is a shortage of domain name space in the current
gTLDs, and if there is, it is not clear that the shortage is serious. While
clearly a great many obvious domain names are already registered by someone, they
are not necessarily being used. Indeed, a great many such registrations are
for the sole purpose of domain name resale (for a profit). So many domain names
that may appear to be unavailable, are in fact available for purchase from domain
name arbitrageurs or distributors. Moreover, there are still a great many "good"
domain names to be had -- just because many of the one word generic domain names
have already been registered does not mean there are no more possible names -- the
various available combinations of words and word elements are still many.
a go slow approach is appropriate, as recommended by the DNSO. No more than
10 new gTLDs should be added, and preferably no more than 5 or 6, with only 1 or
2 completely unrestricted. To the extent that there is a genuine need for more
domain name space, more than doubling the gTLD space on this round seems unnecessary.
Limiting this round of additions to an approximate doubling of capacity will surely
address the need for space in the short term (and perhaps longer), will allow for
an orderly and not crippling expansion of capacity, will avoid the more extreme problems
for trademark owners, and will allow ICANN and all participants in the internet infrastructure
to learn from the experience before moving on to potentially more ambitious expansions
of gTLDs. Given that this will be the first expansion of the gTLDs since the
"internet boom", there can be little doubt that there will be unanticipated problems,
even with the addition of only a few gTLDs, and even with careful planning.
Let's do this round slowly, so that we can learn from it and implement the learnings
in later rounds of expansion.
Third, the new gTLDs should absolutely be subject
to the UDRP. UDRP principles should apply to all gTLDs, regardless of whether
open or restricted. Of course the nature of a particular gTLD may impact the
outcome of cases in that gTLD, but the UDRP should apply nevertheless. The
UDRP is proving to be an effective tool for resolving the most egregious cybersquatting
problems and intellectual property infringements, while leaving the tougher legal
calls to the courts and/or to negotiation between the parties.
Fourth, how to protect
famous mark owners from another massive and expensive battle with cybersquatters?
This is a very hard question. History has shown us that owners of famous trademarks
are victimized by cybersquatters on a very large scale. The volume is a big
part of the problem. Even the availability of the UDRP does not solve the volume
problem for the famous mark holder, who is often faced with literally hundreds of
cybersquatters. So, some sort of mechanism to prevent squatting on famous marks
is truly desirable, if a fair one can be found.
A list of famous marks will
not work and will not be fair, for reasons already debated. It would be virtually
impossible to develop, particularly on a global basis. Moreover, it would unreasonably
add to the cost and bureaucracy of the domain name infrastructure to ask registrars
to implement such a list.
While the last Sunrise proposal was flawed, I
believe that there is the germ of a good idea there. A sunrise (early registration)
system would relieve the infrastructure of any obligation to "police" domain name
applications. Rather, allowing trademark owners to register their marks as
domain names in advance of others has the advantage of leaving it to the mark owners
to solve their own problem -- it is a self-help remedy -- for those that care, they
can spend the money to register. Moreover, if a sunrise registration period
for mark owners were strictly limited to one domain name registration per mark in
each new gTLD, then there would be no registrations by mark owners designed to prevent
non-infringing registrations that constitute parody or free speech. Nor could
mark owners register variations that might be the legitimate marks of others.
For example, in a new .xyz gTLD, Intel would be permitted sunrise registration of
only intel.xyz and pentium.xyz. Intel would not be permitted additional domain
name registrations based on the Intel or Pentium trademarks (e.g., intelsucks.xyz
or pentiumsucks.xyz) during the sunrise period. Rather, should Intel wish to
register intelsucks.xyz in order to prevent someone else from doing so, it would
have to wait until the general registration period, and operate on a first come,
first served basis just like everyone else. In short, only one sunrise registration
in each gTLD should be permitted per registered mark, and the domain name would have
to be identical to the registered mark.
And, of course, the sunrise period would
have to be available only to marks that are in fact registered, at least in one country,
and perhaps more. It should not be available for registration of generic words
as domains. As these are generic words, another person's registration of them
in new gTLDs would not constitute infringement or cybersquatting, given that there
can be no trademark rights established in generic words, regardless of length or
extent of use.
Moreover, no discount should be granted for sunrise registrations.
If trademark owners want to address the problem of cybersquatting by means of prophylactic
domain name registrations (as opposed to use of the UDRP or reactive pursuit of legal
remedies against squatters), then they should pay the fee.
With the foregoing
restrictions (and perhaps others) on a sunrise registration system for trademark
owners, the protection granted to mark holders is very narrow indeed, and is unlikely
to prevent non-infringing or fair uses of domain names that are not identical to
Some have argued that a sunrise registration period for owners
of registered marks disadvantages trademark owners who do not have registrations
for their marks. However, the goal of this system is to prevent cybersquatting,
and in particular squatting on famous marks. It is unlikely in the extreme
that there are marks that are famous or even reasonably well-known that aren't registered
somewhere. And without relying on the existence of a registration, there would
be no assurance that the claimed "mark" is not generic or otherwise unprotectable.
And finally those that claim trademark rights have equal access to the trademark
offices around the world and can register their marks -- registration does not cost
that much, and is well within the reach of even very small businesses. A single
registration in the U.S. generally costs around US$2000-US$3000, far less if use
of outside legal counsel is minimized or eliminated. Owning a trademark registration
is not an unfair requirement to take advantage of a sunrise period.
first blush, it may seem unfair to give anyone a sunrise period, particularly given
the egalitarian culture of the internet and its administration. This may seem
doubly so when it is the lawyers (vile creatures!) that are pushing the idea.
However, there is a genuine problem here, not only for trademark owners, but for
the participants in the internet infrastructure and for the users of the internet
generally. Cybersquatting is a drain on internet resources, is confusing to
the public, is a very expensive problem for trademark owners, and serves no useful
social purpose. Allowing trademark owners a self-help remedy that is narrowly
crafted and that effectively avoids the problem upfront makes loads of sense.
This is not preferential treatment -- its just the only way to effectively avoid
the problem. Further exploration of a sunrise period is merited.
Counsel, Corporate Marketing
Director, Trademarks & Brands