Dispute Resolution Service
27 March 2001
Comments on Nominet's Proposed new Dispute Resolution Policy and Procedure
These comments are offered from my personal perspective as a WIPO UDRP Panelist, who was involved closely with the debates around the development of the UDRP and its earlier unimplemented antecedent - the ACP (originally proposed as part of the gTLD-MoU process). I also have direct experience of taking action against cybersquatters in the UK and elsewhere and of acting for domain name registrants who have been challenged by others.
Nominet's proposals have much to commend them, but do raise one or two major concerns, which are discussed in more detail below.
Mediation and Appeal
- On the positive side, I believe that Nominet's offer of an initial mediation service free to the disputing parties is to be supported. It is a public spirited initiative which reflects the genuinely non-profit, public-service operation of Nominet. Such an initiative could usefully be adopted by others, but sadly is probably unlikely to be taken up by the various profit-driven registry operations we see elsewhere.
- Nominet's proposal to add the facility for an administrative appeal under its policy rather than forcing parties directly into Court is also a desirable feature. In my view, it is a significant shortcoming that the UDRP presently does not presently contain an internal appeal option. This has provided UDRP detractors with a genuine potential grievance against the UDRP when most other general criticisms have little objective merit.
Conditions for using the Policy
My principal misgivings about Nominet's proposals relate to the provisions of Paragraph 2 of the Policy.
- Nominet proposes to depart from the UDRP requirement that a disputed "domain name is identical or confusingly similar to a trademark or service mark in which the complainant has rights". Nominet's new Policy instead refers [para 2a)(i)] to "rights in respect of a name or mark which is identical or similar to a Domain Name" (my emphasis).
- When the UDPR and ACP were being developed, strong objections were expressed by almost all participants to any suggestion that these administrative policies should in any way attempt to develop or apply their own principles of law. The consistent view was that any dispute policy MUST adhere to ESTABLISHED legal principles and precedents. That is why the UDRP is limited to rights in trademarks/service marks (registered and unregistered) which must be CONFUSINGLY similar to a domain name in dispute. "Names" in general were not included precisely because not every jurisdiction recognises rights in "names" as such.
- In my view, therefore, Nominet would be making a potentially risky departure from recognised practice by seeking to apply the new Policy to "names" in general and to "names" which are also merely "similar" to a Domain name.
- Firstly, in the UK it is questionable what "rights in respect of a name" a complainant may objectively claim other than registered trade mark rights and common law rights established by use and reputation. Unless Nominet intends to define further "rights" contractually (e.g. by establishing, say, a .nom.uk namespace for real [as opposed to legal] persons, in which contractually Nominet might be able to assert that a registrant has "rights" if, say, they use that name legally themselves, and in which Nominet would only allow a named person themselves to have the registration of that name), then it is difficult to see how any such additional "rights" could objectively be claimed to exist.
- Specifically, the UK does NOT have personality rights as such. And indeed, in the case of celebrities, following the ELVIS PRESLEY and PrIncess Diana trade mark decisions, it is arguable that celebrities do not necessarily have any recognisable rights in the UK which could be asserted against registration of the celebrity's name as a domain name. Although emotionally it is appealing to suggest (as was done at a recent presentation on the Nominet Policy) that a Dispute Resolution policy should automatically decide in favour of a celebrity complainant if the celebrity's name has been registered as a domain name by someone else, I question whether there is realistically any enforceable legal right which could justify such a policy.
- In fact, ironically, it seems to me more likely that ordinary individuals who are not famous might have such a right, simply on the basis that if their names are misappropriated by others the potential for use of such names as "instruments of fraud" (after One In A Million) is greater because although the public might not expect a celebrity domain name to be operated by the celebrity (after ELVIS PRESLEY), they might not be so circumspect about ordinary individuals in general. [Well it's a thought;-]
- Secondly, whilst it is elsewhere suggested that fans sites or critical sites may be acceptable as "fair use", the proposed requirement for a "name" to be merely "similar" rather than "confusingly similar" to a domain name in dispute, it seems to me that Nominet is opening the door to more potential disputes, extending the scope to grounds beyond those which would be recognised by the UK Courts. Again, this seems to risk Nominet making up its own "law". The concept of "confusing similarity" is well established and understood in law - mere "similarity" could not be a basis for corresponding action in Court in respect of any enforceable "rights" in "names". It may be reasonable to conclude that "<mark>sucks.co.uk" is not "confusingly similar" to a (Trade) <mark>, but I would suggest that <name>sucks.co.uk would almost always be "similar" to <name>.
- With regard to para 2a)(i), therefore, I would suggest that this is amended to be clear that Nominet is not attempting to create some new "rights" [unless it is intending to define these very much more specifically under contract] and is not attempting to extend the scope for conflicts beyond what would be recognised by the UK Courts. A suitable rewording, therefore might be:
"2a)i. The Complainant has rights, which would be enforceable at law in the UK, in respect of a name or mark which is identical or confusingly similar to a Domain Name; and..."
Bad Faith 1
- In Para 2a(ii), Nominet proposes that the second test a complainant must show is that "the Respondent has registered and/or is using the Domain Name in bad faith". The third test in the UDRP, that a respondent has "no rights or legitimate interest in respect of the domain name" has therefore been omitted. Reportedly, this has been done because of a perception that the two corresponding tests in the UDRP are frequently conflated such that there is no purpose in retaining both. [The Nominet change to "and/or" in 2a(ii) in comparison with "and" in the UDRP Clause 4a(iii) is sensible and realistic having regard to the relevant UK interpretation laid down in One In A Million.]
- At the same time, however, and evidently from a moral perspective (rather than strictly legal view) that if a respondent is adjudged to have registered or used a Domain Name "in bad faith" for the purposes of the Policy that is tantamount to convicting the respondent of a criminal offence, Nominet proposes under Para 2b to specify standards of proof for this test which are simply inconsistent with the standards a UK Court would apply to a domain name dispute in similar circumstances.
- Whilst this proposed requirement is presumably driven by a perception that some UDRP decisions have too readily found 'bad faith', in fact I would suggest that has not commonly been the case. Rather, as Jeremy Philips has observed, when one actually looks at the facts of the cases it is surprising that more decisions have not gone against registrants. For a realistic appraisal of the UDRP, I would recommend the recent review by Dawn Osborne "ICANN dispute resolution - A resounding success!" (see http://www.iprights.com/internet/do1000.htm). This article is based on a review of the actual facts and not just a superficial statistical count of decisions (of the sort which has elsewhere been used as a basis for drawing spurious conclusions purely for the purpose of condemning the process altogether).
- Instead of attempting unilaterally to alter the civil standards of proof, I think that it would be more realistic for Nominet to address evidential concerns by applying more rigorous standards to the form and content of evidence. It is in this area, in my experience, where Complainants (and Respondents) can be criticised for making claims without proper support which would not be acceptable in, for example, opposition proceedings before the UK Registry or Community Trade Marks Office. I have certainly seen what can only be described as sloppy complaints which make assertions about rights and claim use and reputation with no effort to support such contentions with anything like the form and content of evidence which would be expected in opposition proceedings.
Deficiencies have included:
- failure to identify the form in which a mark is actually used and/or registered (e.g. logo/ word/ combination etc.);
- failure to show registrations are in force;
- failure to show how Complainant derives rights in a registration or under a mark (e.g. registrations in name of party other than Complainant);
- failure to acknowledge marks under challenge (e.g. existence of opposition);
- failure to support claims to use and reputation (e.g. no information on relevant promotion/sales/turnover/brand recognition etc.)
- failure to support claims to common law rights in similar fashion.
- The UK Trade Marks Registry and OHIM (Community Trade Marks Office) both have guidelines for evidence which Nominet could usefully incorporate expressly into its Policy (i.e. by requiring evidence to be presented in the form and with the substantive content which would be required in trade mark opposition proceedings (which are probably the closest type of proceedings to most domain name disputes). Nominet has got it right in adopting the conventional approach to evidence in 3 rounds (Complaint; Response; Reply to Response), provided Nominet also allows an Expert to call for additional submissions if the Expert wishes. There is no sense in Nominet attempting to set up its own rules of evidence when there is already a well-established framework which works. The framework provided, for example, by Rules 13 and 14 of The Trade Marks Rules 2000, would appear to lend themselves readily to appropriate adaptation.
- The Courts would apply the normal civil standards for evidence and proof to domain name disputes of this sort. Consequently, if Nominet is going to try to require higher standards proof, then we can expect that Complainants will simply be advised to go straight to Court - the costs of issuing a Claim and the proposed £750 Nominet charges are comparable, and a Court case has the benefits that:
- a Court WILL apply the normal civil standards of proof and WILL have due regard to the One In A Million decisions;
- if a registrant fails to respond, a Court can be expected to be more ready to enter a default judgement;
- a losing registrant can be stuffed with a costs order; and
- Nominet will actually have to execute a Court Order [perhaps I have missed it, but I could not see any express statement in the proposed Policy or Procedure to the effect that Nominet would necessarily even put an Expert's decision into effect.]
Bad Faith 2
- The One In A Million appeal judgement addresses issues of trade mark infringement, passing off and capacity of a domain name to be "an instrument of fraud.
In general terms:
"In my view there can be discerned from the cases a jurisdiction to grant injunctive relief where a defendant is equipped with or is intending to equip another with an instrument of fraud. Whether any name is an instrument of fraud will depend upon all the circumstances. A name which will, by reason of its similarity to the name of another, inherently lead to passing-off is such an instrument. If it would not inherently lead to passing-off, it does not follow that it is not an instrument of fraud. The court should consider the similarity of the names, the intention of the defendant, the type of trade and all the surrounding circumstances. If it be the intention of the defendant to appropriate the goodwill of another or enable others to do so, I can see no reason why the court should not infer that it will happen, even if there is a possibility that such an appropriation would not take place. If, taking all the circumstances into account the court should conclude that the name was produced to enable passing-off, is adapted to be used for passing-off and, if used, is likely to be fraudulently used, an injunction will be appropriate.
It follows that a court will intervene by way of injunction in passing-off cases in three types of case. First, where there is passing-off established or it is threatened. Second, where the defendant is a joint tortfeasor with another in passing-off either actual or threatened. Third, where the defendant has equipped himself with or intends to equip another with an instrument of fraud. This third type is probably mere quia timet action."
The Court of Appeal did not apply criminal standards of proof to reach those conclusions. The fact that the domain names in the defendants hands could be considered as potential "instruments of fraud" did not mean that the defendants were being accused or found guilty of the criminal offence of fraud. The case was a civil case, tried on civil standards. Indeed it is notable that the Court found no need to refer expressly to "bad faith" as such.
- Perhaps the problem with the UDRP and Nominet Policy is that the term "bad faith" is seen as conveying moral condemnation beyond its legal intent and I wonder if in fact substitution of some alternative terminology such as (say) "use or registration without applicable authority or not otherwise in accordance with legitimate practices" or simply "use or registration not being bona fide" (the Latin is perhaps less emotive than "bad faith") [these are just suggestions, but if others are finding the words carrying heavy moral baggage then consideration could be given to adopting alternative wording.] From my own perspective, I wonder if an acceptable alternative might be adapted from the second UDRP test which Nominet proposes to drop: e.g. simply:
"The Respondent has no rights or bona-fide or legitimate interest in the Domain Name"
The fact that the Domain Name is (or is not) being used would then be relevant to the existence of a "bona-fide or legitimate interest", so from a legal, analytical perspective the test would be essentially the same, but without an express finding of use or registration "in bad-faith", which seems to cause such emotional overreaction (as a Panelist I have been threatened with suit for libel for deciding, for the purposes of the UDRP, that a registration had been made in bad faith!)
- It has also been said that Nominet takes the view that "bad faith" is not equivalent to the absence of good faith. In fact, UK and European Community Trade Mark law have well established views on what constitutes "bad faith" (see e.g. Kerly - Law of Trade Marks and Trade Names 13th Ed. ss.7-193 to 7-230), which suggest that, if accurately stated, Nominet's view is not legally justified. In Gromax (1999RPC367) Lindsay J. stated that "bad faith" certainly "includes dishonesty" and "some dealings which fall short of the standards of acceptable commercial behaviour observed by reasonable and experienced men in the particular area being examined". G.Hobbs QC in Demon Ale (2000RPC345), referring to the dicta of Nicholls L.J.in Royal Brunei Airlines 2AC378 said "a finding of bad faith may be fully justified even in a case where the applicant sees nothing wrong in his own behaviour." At the Community level, the OHIM Cancellation Division in Trillium (Case C53447/1) stated unequivocally: "..bad faith is the opposite of good faith, generally implying or involving, but not limited to, actual or constructive fraud, or a design to mislead or deceive another or any other sinister motive." As Kerly concludes, I believe correctly, these various decisions suggest that the appropriate test for "bad faith" must be substantially an objective one based on the facts, and certainly cannot depend significantly on the state of mind (e.g. perceived innocence or alleged lack of knowledge) of the applicant, although in a finely balanced case, such subjective elements may make the difference.
- In conclusion, therefore, I would suggest that Nominet should delete proposed para 2.b from the Policy. I would suggest that Nominet could achieve the desired effect of tightening up on decision making by requiring the evidence to meet the CPR standards as applied by the Trade Marks Registry to evidence in Trade Marks oppositions. If explicit reference to "bad faith" is desired, then, consistent with Kerly and the TM Registry Practice, I would suggest including a statement that an allegation of "bad faith" must be particularised with evidence to support such a claim.
- Under para 16 d. of the draft Procedure (Rules), Nominet proposes a "three strikes and you're out" provision by which it would exclude use of the Policy by any Complainant adjudged to have bought three complaints in "bad faith", for example, "in an attempt at Reverse Domain Name Hijacking". In my view this one-sided provision has serious shortcomings for a number of reasons.
- it IS one-sided - there is no corresponding provision for penalising abusive registrants caught out three times. It therefore gives the impression (along with para 2b. of the Policy) that Nominet's DRS is perhaps not as balanced as it would like to claim. I do not accept the argument that it would be impossible for Nominet to refuse or impose conditions on further registrations by abusive registrants - it may be easy for them to get around simple exclusions, but then any such circumvention would be likely to be prima facie evidence of bad faith in subsequent proceedings, for example. In fact, I am not personally in favour of bans, as such, but believe that the appropriate sanction in cases of abuse (by registrant or complainant) ought to be in terms of costs (see below for suggestions).
- "Reverse Domain Name Hijacking" does not appear to be defined either in the Policy or the Procedure - an appropriate definition as in the UDRP Rules needs to be added.
- Just as for showing "bad faith" registration, it should be made clear that such an allegation against a Complainant by a Respondent must be particularised with supporting evidence (see comments on evidence above).
- just as Nominet is correct in noting that the number of domain name disputes is small (from my own experience, the real number will certainly be at least two or three times greater than those which actually come to Nominet's attention), so the number of genuinely abusive Complaints is probably a couple of orders of magnitude smaller still.
- Personally, rather than attempting to treat rogue Complainants as vexatious litigants and unilaterally excluding them from any use of the Policy, it seems to me that the proper penalty should be in costs. I would suggest that if a Complainant is found to have brought "bad faith" complaint on even just TWO occasions then Nominet should simply require that the fee for subsequent Complaints from the same Complainant is multiplied by the number of "bad faith" judgements against it (i.e. two bad faith judgements and the next complaint costs double, three and it costs triple etc.) This would discourage abuse while not excluding anyone from seeking a remedy if they actually do have a case.
- Correspondingly, for abusive registrants found to have made multiple bad faith registrations in (say) three or more separate cases, the Rules could require the Respondent to pay a sum (say £250) to Nominet before any Response would be considered in subsequent Complaint proceedings, with Nominet repaying this if the Registrant is successful or passing it to the Complainant if not. Despite the protestations of general innocence from the registrant lobby, I have no doubt that other practitioners in the field will confirm my own experience that there are still plenty of abusive registrants out there who have adapted to the UDRP simply by tailoring their demands to a sum which is slightly below the costs of a UDRP filing. I have no doubt that in the UK these people will be asking Complainants for £500 to buy them off if Nominet's charge for using the Policy is to be £750. It would be nive to assume otherwise. To ask a Respondent to front up with £250 to defend subsequent cases after three decisions against them does not seem to me to be unreasonable in such circumstances.
Declaratory Judgement option for threatened Registrant
- I would also support the suggestion by Matthew Rippon of Prettys that consideration should be given to providing a procedure for a domain name registrant to pre-empt a Complaint by seeking some sort of Declaratory Judgement that their use/registration of an domain name does not fall within the Policy. I don't think this is as simple a matter as Matthew has suggested - it would probably entail a reversal of most provisions and should certainly require a registrant to have grounds in the form of a threat from a potential complainant so that the action might be akin to a threats action under the Trade Marks Act, for example (as in the Prince.com litigation). However, it is certainly an option which merits further attention.
- One of the most significant difficulties faced by any Complainant is verifying a Registrant's contact details and identifying the extent to which a registrant may be involved in making systematic abusive registrations. The WHOIS facility is essential for this purpose. Nominet's WHOIS is one of the least helpful, in that it does not even provide an address for service and does not allow identification of all the domain names registered by a given registrant. (Contrast the information readily available from the Trade Marks Registry online databases and from the Companies Register on registered addresses and directors' names.) Nominet has argued that Data Protection provisions restrict its ability to provide such information. As anyone who has used an email address is aware, mailing and contact list details are regularly passed on such that Data Protection provisions do not pose an insurmountable obstacle. At the very least, for domain names in any namespace which purports to be commercial (e.g. .co.uk) then it should be possible to have a registrant's name and address for service and a list of all names held by that registrant. Nominet can act under the copyright laws relating to Database rights to prevent unfair and abusive extraction of WHOIS information for spamming, for example, whilst allowing access for legitimate enquiry. Even if Nominet is reluctant to provide such information publicly then it should provide that information on request. If Complainants are forced to seek Norwich Pharmacal orders to get the necessary information from Nominet then they are likely to continue in Court and the object of having the Policy will be defeated anyway.
- I hope these comments, even if they have necessarily focussed on some areas of criticism, are also helpful. Overall, I do feel Nominet is moving in the right direction and would commend it for this. I would be happy to expand on these comments if those responsible for developing Nominet's Policy would like to discuss any of the issues I have raised in more detail.