Comments regarding UDP

[Carl Oppedahl <carl@oppedahl.com>]

On  September 29, 1999, ICANN posted to its web site two documents:
"Uniform Domain Name Dispute Resolution Policy (the 'Policy')" and "Rules
for Uniform Domain Name Dispute Resolution Policy (the 'Rules')" and
invited public comment regarding the documents.

It would be all too easy to limit one's comments to the documents
themselves, and indeed there are many things wrong with the documents
themselves.  But as one who has been involved with domain name disputes for
many years, probably longer than anyone who participated in the drafting of
the Policy and Rules, I see fundamental problems with ICANN's actions.  The
ICANN process leading to this Policy and Rules betrays ICANN's fiduciary
duty to the Internet community.  The process was skewed from its beginning
to favor the interests of the trademark lobby and gave no meaningful voice
to the domain name owners of the Internet.  The past four years have seen
numerous instances of reverse domain name hijacking, yet the ICANN process
included no one who had ever been a target of reverse domain name
hijacking, and included no lawyer who had ever represented a target of
reverse domain name hijacking in court.  The ICANN process was stacked from
its outset with lawyers who have built their careers on expansive
interpretations of trademark law.  It is, perhaps, unsurprising for these
reasons alone that the Policy and Rules are extraordinarily biased against
domain name owners and in favor of trademark challengers including reverse
domain name hijackers.

A first test, and the best test, of the legitimacy and fairness of ICANN's
actions would be to see whether this Policy and Rules somehow respond to
the expressed wishes of the Internet community.  The answer is, they do
not.  Other than the trademark lobby, no individual or group within the
Internet community has asked for a uniform dispute resolution policy at
all, let alone the Policy and Rules being discussed now.  

Another test of the legitimacy and fairness of ICANN's actions would be to
see whether this Policy and Rules somehow respond to the expressed wishes
of the domain name registrars.  The answer is that they do not.  The
approximately twenty domain name registrars who drafted the policy and
rules which were the starting point for ICANN's deliberations reached
consensus on a "Model Dispute Resolution Policy for Voluntary Adoption".
The document prepared by this group of twenty registrars emphasized that
the policy should be "voluntary," using the term twice in the first twenty
words of the document.  It was clear from the registrars' Model Dispute
Resolution Policy for Voluntary Adoption that the registrars intended that
each registrar be permitted to make its own decision whether to adopt the
"voluntary" policy or some other policy.

If the observed behavior of the present-day registrars had all shown
adoption of some particular policy, this might suggest that there was a
consensus among them for a uniform policy.  Instead, we see for example the
policy adopted by NameSecure.com which limits its role in domain name
disputes to providing contact information to the parties.  This is the very
domain name dispute policy drafted by Jon Postel in RFC 1591 in 1994, a
policy followed to this day by many registrars around the world.  

One of the most striking signals that the ICANN Policy and Rules lack
legitimacy is the set of ground rules announced by ICANN for the almost
nonexistent comment period provided by ICANN, a fourteen-day period that
ends today.   ICANN has said that the only comments which ICANN will
consider are at this time are those directed to the narrow question of
whether the Small Drafting Committee did or did not fulfill the mandates
given to the Committee by the ICANN board in Santiago, Chile.  In saying
this, ICANN has explicitly plugged its ears to any substantive comments by
the Internet community as to the global questions such as whether the
policy needs to be "uniform" (that is, forced upon all registrars and
through them, upon all of the six million domain name owners within COM,
NET, and ORG) and the question whether it is even needed at all.  After
all, NSI has reported a decline in the number of domain name disputes in
recent years.

Three years ago I wrote a law review article "Analysis and Suggestions
Regarding NSI Domain Name Trademark Dispute Policy," 7 Fordham Intellectual
Property, Media & Entertainment Law Journal 73, an article pointing out
flaws in NSI's policy and offering pages of detailed suggestions to those
who would try to draft a successor policy.  I provided copies of this
article to all of the ICANN board members when they were appointed, and
provided copies of this article to all of the members of the ICANN Small
Drafting Committee.  A close reading of the present Policy and Rules shows
every recommendation in that law review article to have been ignored.

During the weeks in which the Small Drafting Committee deliberated, many
members of the Internet community (I among them) provided comments and
suggestions to the Committee to try to assist it in addressing the problem
of reverse domain name hijacking, and in removing the extreme pro-trademark
biases.  Not one of these suggestions was adopted by the Committee.  Every
change made by the Committee during its secret deliberations was in the
pro-trademark direction.

As mentioned above, the ICANN board stated in Santiago that the standard of
review of the Committee's results would not be whether the Committee's
results were fair or reasonable or balanced, but merely whether the
Committee had fulfilled the small handful of mandates from the ICANN board.
 If one holds the hoop low enough, almost anyone can jump through it.
Remarkably, the Small Drafting Committee failed to jump through even this
hoop, or to say this another way, the Policy and Rules flunk even the
modest test set by the ICANN board in Santiago.  The ICANN board directed
the Small Drafting Committee to achieve the following results:

          4. There should be a general parity between the appeal rights of
complainants and domain name holders. 

          5. The dispute policy should seek to define and minimize reverse
domain name hijacking.

Parity of appeal rights.  Under the Policy and Rules, a domain name owner
whose case was wrongly decided by the tribunal faces the almost impossible
task of filing suit within a mere ten days.  During this brief time, the
domain name owner must interview candidate counsel, select candidate
counsel, ask candidate counsel to do a conflict check, wait for the results
of the conflict check, and then perhaps repeat the process if the conflict
check fails.  Once counsel are selected, the site of the blocking lawsuit
must be selected.  Once the site is chosen, the process must be repeated to
identify, select, conflict-check, and retain local counsel in the proposed
forum.  There, too, if the conflict check fails the domain name owner must
start over in the selection process.  Then counsel must learn all the facts
of the case and prepare court papers for filing.  These must, of course, be
reviewed by the domain name owner for approval.  The case must be filed,
and a court-stampted copy of the papers obtained (at least one day before
the end of the ten-day period) and sent by overnight courier to the ICANN
tribunal.

All of this must be accomplished within ten days of the tribunal's decision.  

>From the above discussion it should be apparent that ten days is woefully
inadequate.

When NSI announced its universally criticized domain name dispute policy in
1995, one of the first flaws immediately identified by the Internet
community was the fourteen-day period during which the domain name owner
would have to file suit to block NSI's carrying out its decision regarding
the domain name.  Even NSI, not known for doing fair or reasonable things
for its domain name owner customers, was shamed into lengthening this
period, first to thirty days and then to thirty-seven days.  I can speak
from person experience, having counseled many innocent domain name owners
who wished to appeal NSI's adverse decisions within the 14- or 30- or
37-day period.  My experience is that it takes many innocent domain name
owners at least a week simply to find competent counsel to handle such a
dispute, let alone initiate the conflict check and retain counsel.  

Having said all this, recall that the question presented is whether there
is parity of appeal rights as between domain name owners and challengers.
And the answer is that there is no parity, but instead there is a profound
disparity.  While the domain name owner stuck with a wrong decision by an
ICANN tribunal is required to sue within ten days, the challenger stuck
with a wrong decision by an ICANN tribunal has no ten-day limit.  in fact,
a challenger who is unhappy with an ICANN decision can wait a year or two
years or more and simply go to court to get a second bite at the apple. 

The Policy and Rules drafted by the Small Drafting Committee and posted by
ICANN on September 29 fail to even pretend to provide parity of appeal
rights.  The Small Drafting Committee failed to jump through even this low
and easy hoop.

Another of the mandates from the ICANN board to the Small Drafting
Committee was that "the dispute policy should seek to define and minimize
reverse domain name hijacking."  In fact the Policy lacks a definition of
reverse domain name hijacking.  (There is a watery definition in the Rules,
but none in the Policy where the ICANN board required it.)  More
importantly, nothing in the Policy or in the Rules responds to the mandate
to "minimize" reverse domain name hijacking.  The only changes to the
Policy or Rules made by the Committee during its deliberations were changes
that actively promote reverse domain name hijacking.

As was explained to the Small Drafting Committee by several commenters,
there are only two possible ways in which the designer of such a Policy and
Rules could minimize reverse domain name hijacking.  One way would be to
provide a "gatekeeper" mechanism that would provide bright-line rules for
excluding certain cases so as to minimize reverse domain name hijacking.
The remaining way would be to provide some sort of end-of-the-case result
that would compensate the target of reverse domain name hijacking or
otherwise deter other would-be reverse domain name hijackers from carrying
out their plans against innocent domain name owners.  Regrettably, the
Committee failed to do either of these things.  Instead, the Policy and
Rules drafted by the Committee actively extend a welcome mat to reverse
domain name hijackers, explicitly eschewing any gatekeeper mechanism and
promising in advance not to assess any monetary penalty against a
challenger no matter how frivolous or groundless the challenge turns out to be.

What should be done:

A.  In compliance with the wishes of the twenty registrars who drafted the
policy later used as a "starting point" by ICANN, the policy should be
restored to "voluntary" status.  In this way, the wishes of the Internet
community will be best served, because registrars will be able to select
their own dispute policies as they do now.  If (as against all evidence)
there are domain name owners who actually want a policy such as that
published September 29, 1999 by ICANN, then the market will presumably
provide it because some registrar will adopt the "voluntary" policy.  If
(as the evidence seems to indicate) the Internet community prefers the RFC
1591 rule drafted by Jon Postel, then registrars will adopt that policy (as
NameSecure.com has done).  One doesn't need ICANN, actually, to find out
and fulfill the wishes of the Internet community regarding domain name
dispute policies.  The market, if allowed to continue to do so, will
fulfull the wishes of the Internet community.  There is simply no need for
"central planning" when the goal is simply to provide what the Internet
community wants.

B.  Meaningful and effective provisions must be added to the Policy and
Rules to minimize reverse domain name hijacking.  One easily implemented
and objective standard would be to make the ICANN tribunals available only
to those complainants whose trademarks are unique and coined.  NSI's
failure to apply this simple entrance criterion has led to numerous highly
visible failures of the NSI policy, among them the regis.com case, the
roadrunner.com case, the prince.com case, the perfection.com case, the
elegance.com case, the war.com case, the cds.com case, the earth.com case,
and the clue.com case.  Despite advice from many commenters, the Small
Drafting Committee failed to include entrance criteria that would have kept
any of these cases out of the ICANN tribunal.  

Another easily implemented and objective standard would be to deny the
availability of the ICANN tribunal in cases where the challenger fails to
do anything until after the domain name owner has been using the domain
name for a long time.  Instead, the Policy and Rules impose no time limit
at all on the challenger.  This leads to the remarkable result that a
would-be challenger who considered suing in court but discovered that had
allowed a statute of limitations period to pass would then proceed to the
ICANN tribunal precisely because it imposes no time limit or statute of
limitations.

Yet another easily implemented and objective standard would be to require
the would-be challenger to post bond before proceeding with the ICANN
challenge.  If the domain name owner found it necessary to appeal, and if
the appeal resulted in a reversal, the bond would be paid over to the
domain name owner to compensate the domain name owner for the costs of the
defense and appeal.  Would-be reverse domain name hijackers would be
deterred from hijacking by the risk of loss of the bond.

Other changes were proposed by commenters to the Small Drafting Committee
which might have helped comply with the mandate to "minimize reverse domain
name hijacking", but all of them were apparently ignored, as the Policy and
Rules contain nothing to minimize reverse domain name hijacking.

For the reasons set forth in the preceding paragraphs, the Policy and Rules
don't even manage to jump through the low hoop set for them by ICANN.
There are, however, other betrayals of the Internet community in the Policy
and Rules, some of which will be described below.

Secrecy.  The Policy, as drafted, permits the ICANN tribunals to slap
secrecy orders on their decisions in their sole descretion.  This repeats
one of the flaws in the NSI policy, namely NSI's decision to keep its
decisions secret.  The Policy and Rules should provide that all tribunal
decisions be published.  There is nothing quite like the disinfectant of
bright sunlight to minimize bias, unfairness, or error.

Location of the blocking lawsuit.  The Policy and Rules, as drafted, impose
narrow restrictions on where the blocking lawsuit has to be filed if an
incorrect decision by an ICANN tribunal is to be blocked and appealed.  The
restrictions should be removed.

Transfers.  The universally criticized NSI policy at least managed to avoid
transferring domain names to undeserving parties, but limited itself to
placing domain names "on hold".  In contrast, the present Policy and Rules
explicitly empower the ICANN tribunal to select a new owner for a contested
domain name and to order its transfer.

Yet as history has shown, it is commonplace for a reverse domain name
hijacker to assert a mark used by dozens or even hundreds of companies
(e.g. the clue.com, juno.com, prince.com, perfection.com, elegance.com, and
earth.com cases).  In such a case, there is no sound reason to award
ownership of the coveted domain name to the challenger whose only
credential is that of having won the race to the courthouse.  When the
Prince Sporting Goods company brought its NSI challenge against the
computer training services company that owned prince.com, there would have
been no reason to think that its claim to the coveted domain name was any
stronger than that of the hundreds of other companies named "Prince".  Yet
ICANN is prepared to empower its tribunals to select one trademark owner
over many others as the supposedly deserving recipient of a domain name.
It's wrong to do so.  Tribunal-ordered transfers of domain names should
happen only in the (rare) cases where the asserted trademark is
demonstrably unique and coined.  Denying transfer-the-domain-name relief to
those who are "one among many" users of a dictionary-word trademark would
go a long way toward minimizing reverse domain name hijacking.

I wrote about this "one among many" problem in a law review article
"Remedies in Domain Name Lawsuits: How is a Domain Name Like a Cow?", 15
John Marshall Journal of Computer & Information Law 437 (1997).  Copies of
this article were personally provided to each member of the ICANN board and
to each member of the Small Drafting Committee.  Not one of the suggestions
offered in that law review article can be found anywhere in the Policy or
Rules.

Many others have criticized the ICANN Policy and Rules as I have in this
document.  My experiences with the ICANN process prompt me to express
skepticism that any of these criticisms will be paid any attention to by
ICANN.  My prediction is that ICANN will make no changes to the Policy or
Rules that remove pro-trademark bias, and that if ICANN makes any changes
at all to the September 29, 1999 version, the changes will be to kowtow
further to the trademark interests.  If that's how things go then ICANN
will have deeply betrayed its constituents, the domain name owners of the
Internet community.

Carl Oppedahl