Categories
UDRP

Domain hijacking and bias in UDRP panelists

Statistically, the UDRP panelists are predisposed to find for complainants. Some panelists find for the complainant in 87% of the cases. Many of these panelists have decided hundreds of cases. Trademark owners filed an all-time record 3,036 cases under the Uniform Domain Name Dispute Resolution Policy (UDRP) with WIPO in 2016, an increase of 10% over the previous year, so the stats on panelist decisions have a high degree of confidence.
Even the two so-called “fair” panelists claim that their decisions fall 50-50 for and against Complainants, according to their statements at an INTA 17 panel discussion in Barcelona. These two facts put together are shocking to many domain name owners. How can most panelists usually find for complainants while the fair panelists do so only 50% of the time? There must be an unfair bias in the process.
The UDRP was developed to address the “deliberate, bad faith registration as domain names of well-known and other trademarks” (WIPO Final Report, Par. 23 (1999)).
However, there is a growing and unfortunate trend of companies seeking to acquire domain names wrongfully by asserting spurious cybersquatting claims as a “plan b” when their purchase negotiations do not lead to a price they are willing to accept. This behavior is called reverse domain name hijacking (“RDNH”) or just domain hijacking.
While an RDNH declaration by a WIPO panel has no monetary implications in of itself, it puts WIPO on notice that a Complainant filed an abusive complaint, and it is effective in deterring abusive complaints. US-based domain name owners may sue in a District Federal Court to collect on damages of up to $100,000 per domain name that was the subject of a complaint. This remedy is not limited to cases where WIPO panelists declared an RDNH.
UDRP watchers often note when they can’t believe that a decision wasn’t RDNH, when they believe it should have been.

Attempted Domain Hijacking

For example, in the recent WIPO decision over Soti.com, Andrew Alleman wrote a post titled I can’t believe it’s not RDNH (again), saying, “It looks like a clear case of reverse domain name hijacking to me. But panelist Nick J. Gardner bails out SOTI Inc. out.”
I believe that Andrew Alleman’s assessment is correct for the following reasons:
Firstly, the Complainant applied for their SOTI trademark in Canada on 2009-12-04 and claimed use in Canada since June 27, 1995, which is the day they incorporated. They claimed to have used their SOTI trademark on all of the following services since the day they incorporated:

(1) Computer software, namely for software for the synchronization, management, tracking and security of data transmitted over wireless data networks between a stationary unit and multiple mobile computers or electronic devices in the field of mobile device management.

It was rare for a 1995-era software business to hit the ground running with a fully functional website and complex computer software on day one. It should be noted that Canada does not require trademark applicants to make any verified statement of accuracy under penalty of perjury, so the 1995 date of first use should not be accepted at face value. And while wireless data networks existed in 1995, it was rare to find anyone using it in Canada.
Secondly, the Complainant’s trademark is registered only in Canada. SOTI Inc. is an Ontario, Canada, based corporation. Its rights should not have any impact on the Respondent’s USA-based usage of a .com domain name and in a UDRP. Canada has its own CDRP dispute system for .ca domain names. UDRP’s are not for arbitrating priority of global trademark rights.
Thirdly, the domain name owner (or Respondent) registered their soti .com domain name under its corporate name Strategic Outsource Technologies, Inc. on 1998-01-19; that is 11 years before the Complainant applied for its trademark. The Complainant failed to explain how the domain owner would have known about and targeted the Complainant in January 1998.
And fourthly, SOTI is an acronym for the domain name owner’s long corporate name, which is an obvious legitimate interest in a domain name.
Objectively, the Complaint failed to prove at least two of the three UDRP requirements.
While the panelist, Nick J. Gardner, denied the complaint, he failed to declare an RDNH. It may be that the panelist was biased in favor of the Complainant because its website makes it appear to have a successful business while the Respondent’s soti .com website appears to be less successful. This appears to be a trend among UDRP decisions. The Complainant’s soti .net website makes it appear that they have a successful business.
The Panelist Considered RDNH
Mr. Gardner wrote in his decision:

Several circumstances of this case have led the Panel to consider whether the filing of the Complaint constitutes “using the Policy in bad faith to attempt to deprive a registered domain name holder of a domain name” see the Rules, paragraph 1, definition of Reverse Domain Name Hijacking (“RDNH”). …

It must have been apparent to the Complainant that the Respondent would say he was operating a bona fide business called “Strategic Outsource Technologies Inc.”, and that he independently chose the Disputed Domain Name as corresponding to an acronym for that business. However, it seems to the Panel that the Complainant’s failure to address this issue may be as a result of a misapprehension on the Complainant’s part as to what it has to show. The Panel also has noted that the nature and content of Respondent’s website (as described above) does appear to be rather odd, and provides at least some support for the Complainant concluding there was a case to be answered. The Panel also notes the Respondent has not provided any further statement responding to the Complainant’s Supplemental Statement. Taking all of this together the Panel does not think it appropriate to conclude this case was brought in bad faith.

In all the circumstances the Panel, therefore, declines to find that the Complaint was brought in bad faith and constitutes an abuse of the administrative proceeding.

However, the panelist took a limited view of the criteria for RDNH, when he wrote:

Paragraph 15(e) of the Policy provides that if after considering the submissions the Panel finds that the complaint was brought in bad faith, for example in an attempt at RDNH or was brought primarily to harass the domain name holder, the Panel shall declare in its decision that the complaint was brought in bad faith and constitutes an abuse of the administrative proceeding.

Specifically, the panelist failed to consider other circumstances for finding RDNH. Reasons articulated by panels for finding RDNH include:

(i) facts which demonstrate that the complainant knew it could not succeed as to any of the required three elements – such as the complainant’s lack of relevant trademark rights, clear knowledge of respondent rights or legitimate interests, or clear knowledge of a lack of respondent bad faith (see generally section 3.8) such as registration of the disputed domain name well before the complainant acquired trademark rights,

(ii) facts which demonstrate that the complainant clearly ought to have known it could not succeed under any fair interpretation of facts reasonably available prior to the filing of the complaint, including relevant facts on the website at the disputed domain name or readily available public sources such as the WhoIs database,

(iii) unreasonably ignoring established Policy precedent notably as captured in this WIPO Overview – except in limited circumstances which prima facie justify advancing an alternative legal argument,

(iv) the provision of false evidence, or otherwise attempting to mislead the panel,

(v) the provision of intentionally incomplete material evidence – often clarified by the respondent,

(vi) the complainant’s failure to disclose that a case is a UDRP refiling,

(vii) filing the complaint after an unsuccessful attempt to acquire the disputed domain name from the respondent without a plausible legal basis,

(viii) basing a complaint on only the barest of allegations without any supporting evidence.

http://www.wipo.int/amc/en/domains/search/overview3.0/#item416

Bias and Domain Hijacking

At another INTA 17 session in Barcelona, there was a really good interactive talk on bias. Bias is insidious. It operates in an inconspicuous or seemingly harmless way but bias actually has a grave effect. Most judges strive to be unbiased but often they are not.
My takeaway from the session was that bias is instinctual and more common amongst those with experience or an agenda and a checklist. If you see a grizzly bear, you don’t carefully evaluate the situation; you grab your companion and you run away immediately. However, the bear may have been busy eating fish and had no intention of harming you or your companion.
UDRP panelists often use a checklist when evaluating a complaint and response. Since panelists’ focus on their checklists, they usually miss facts that should cause complaints to be denied outright.
Complaints should be screened by an AI that looks for limitations in their allegations. If certain facts are not met, then the complaint should be denied outright and RDNH should automatically be declared.
An AI could think about a complaint in a scientific way. It could find holes in allegations and bounce complaints. For example, in chemistry, a limiting reagent (or limiting reactant) in a chemical reaction is the substance that is totally consumed when the chemical reaction is complete. The amount of product formed is limited by this reagent, since the reaction cannot continue without it.
The limiting reagents are the following three criteria for a successful complaint are enumerated in paragraph 4(a) of the Policy:

(i) the disputed domain name is identical or confusingly similar to a trademark or service mark in which the Complainant has rights;

(ii) the Respondent has no rights or legitimate interests in respect of the disputed domain name; and

(iii) the disputed domain name has been registered and is being used in bad faith.

To succeed, the Complainant must demonstrate that all of three elements in paragraph 4(a) of the Policy to be satisfied.
Panelists should never make judgment calls in evaluating complaints or overemphasize certain aspects of a Respondent’s use of a domain name or anything derogatory about a Respondent. Either a complaint objectively states a case or it doesn’t.
Now that WIPO has published its Jurisprudential Overview 3.0, an AI can be taught to check whether a complaint states sufficient objective facts to permit it to move on for evaluation by a panelist. It will be a challenge for the AI to not learn and hold the same biases that have plagued WIPO decisions to date, if it were to only review all of the decisions and establish decision trees based on what it has been previously written.
The Jurisprudential Overview 3.0 provides: where a domain name has been registered before a Complainant has acquired trademark rights, only in exceptional cases would a complainant be able to prove a respondent’s bad faith. Since trademarks are territorial and not global, I would add that the trademark rights have to match those of the TLD extension.
So only an EU trademark registration would have any bearing on a .eu domain name. And only a USPTO trademark registration would have any bearing on a .com, .net, or .org domain name. Since .ca domains are not part of the UDRP system, and since most Canadians expect Canadian websites to use .ca domain names, Canadian trademarks should have no bearing on any other TLD.
So if the USPTO has determined that a Complainant’s trademark application to register a word or phrase that comprises the domain name is generic, then the complaint should automatically be bounced as RDNH.
For example, the USPTO determined that the phrase “sesame snaps” is generic with respect to cookie snacks that a Complainant sells. Under no circumstances should a panelist entertain a complaint about sesamesnaps.com brought by that same person. A USPTO determination that a trademark application is not registrable because it is generic is prima facie evidence that the domain is generic. The Complaint should be automatically declared RDNH. But no, Tony Willoughby who claims to be one of the fair panelists, found in favor of the complainant instead. What does Tony do? Does he throw complaints down a set of stairs and make decisions based on whether a complaint falls on an even or an odd step? Or is something else going on? Maybe it is a bias based on his checklist. He missed the limiting reagent in the process.
If facts are in dispute, then panelists should deny the complaint. The UDRP is not a court that permits a person to a full right to be heard and confront his accuser when determining property rights. Some domain names are worth $1000. Others are worth $1,000,000.  It does not make financial sense to appeal every wrong-headed UDRP decision over most domain names. But a UDRP loss has a significant future impact on a domain name owner who also has other valuable domain names.
A UDRP loss is often cited in future UDRP complaints and even in court cases.  Not every domain name owner has an extra $50,000 to throw away at litigating based on the principle of the matter. Sometimes you have to sell other domain names at less than full value to make sure you clear your name and win.
Some wealthy domain owners, like Telepathy Inc., sue for damages under the ACPA whether they win or lose a UDRP, to collect on statutory damages for domain hijacking.

Common Examples of Bias

A bias is sometimes created by providing inadmissible evidence in a proceeding. Lawyers should be sanctioned for doing so and often are in the USA. But not in UDRPs, CDRPs, or in some foreign courts, for example, Canada.
The US Federal Rules of Evidence try to prevent a jury from hearing inadmissible evidence, under Rule 103:
(d) Preventing the Jury from Hearing Inadmissible Evidence. To the extent practicable, the court must conduct a jury trial so that inadmissible evidence is not suggested to the jury by any means.

The judge can foreclose a particular line of testimony and counsel can protect his record without a series of questions before the jury, designed at best to waste time and at worst “to waft into the jury box” the very matter sought to be excluded.

A bias can also be introduced as a single-minded focus. A single-minded focus is not good when you are trying to judge a situation.
When driving on highways, most drivers focus on keeping their distance from other cars or following other cars at a fixed distance. When the lead car hits something, sometimes there is a pileup of cars: a chain reaction happens.
How often have you heard about pedestrians and moped drivers being hit on the roadway? Drivers who stay at the scene usually tell police that they didn’t see the bike. Of course, they didn’t see the bike. They were too busy focused on looking out for other cars and keeping within the speed limit.
Domain hijacking often occurs because there is bias against domain owners who do not have a successful business and can’t afford to hire experienced counsel. It often takes years before entrepreneurs can grow their business to the point where they have significant excess cash flow. Some panelists are biased in favor of successful businesses.

Implicit Bias

Thoughts and feelings are “implicit” if we are unaware of them or mistaken about their nature. We have a bias when, rather than being neutral, we have a preference for (or aversion to) a person or group of people. Thus, we use the term “implicit bias” to describe when we have attitudes towards people or associate stereotypes with them without our conscious knowledge. A fairly commonplace example of this is seen in studies that show that white people will frequently associate criminality with black people without even realizing they’re doing it.
You might only be 1% less likely to find for a particular party due to some bias on a given motion. But that 1% difference over a course of motions and judgment calls may turn into a 40% difference.
Read more about implicit bias at the Perception Institute’s website.
Check back for more updates on how bias develops and may be reduced in UDRPs.