This article is more than 1 year old

Why ICANN's domain dispute rules are flawed: Part I

The history and problems with the policy

The Uniform Dispute Resolution Policy (UDRP) is an exceptionally important Internet process and stands to become all the more important in years to come. It is being adopted by new top-level domains; ICANN and WIPO are pushing it into the wider international arena and country-code TLDs; and the European Union plans to adopt it for the new .eu domain.

On top of that, the draft proposals for the Free Trade Area of the America, recently revealed, include UDRP as the policy for domain disputes across all countries in North, South and Central America.

What is UDRP? Basically, the set of rules that decides who or what has the right to a particular domain name.

However, ever since the first domain disputes cases went through the ICANN system, we have consistently prefixed UDRP with the adjective "flawed". It is time, we figured, to spell out exactly why this set of rules threatens to damage the very fabric of the Internet.


It's important to note that the origin of UDRP was very different to how it has now turned out. ICANN asked WIPO to produce a report on how Internet domain disputes could be quickly, simply and cheaply settled to save the Internet from becoming a huge, litigious no-man's land.

WIPO produced a report, with a range of recommendations, and submitted it to ICANN. There, it was widely discussed and eventually the Uniform Dispute Resolution Policy was formed and approved on 24 October 1999.

The most important aspect to note is that the policy was solely for trademark owners in dispute with non-trademark owners. UDRP specifically refused to rule on personal names, place names, names of pharmaceutical drugs and government organisations. Those names should follow their usual course through the law courts, it was decided, since law varies in different countries.

The full policy was forced onto domain name registrars, which allowed for an effective and fast resolution process across large swathes of the Internet.

Unfortunately, in trying to make the policy as simple and cheap as possible (so ordinary Net punters would not be disadvantaged), several important aspects were overlooked or mishandled. This, coupled with over-zealous arbitrators, has turned UDRP into a deeply flawed, corporate-friendly law that is precariously riding upon its own errors of judgement.

So what the hell are UDRP's flaws?


To our mind there are ten basic problems with the policy:

  1. It does not spell out who the burden of proof falls on
  2. It does not say what evidence is required to prove innocence or guilt
  3. It imposes impossible time limits
  4. A complainant does not need to contact a domain name owner before going to an arbitrator
  5. The complainant decides which arbitrator to go to
  6. There is no appeals process
  7. Arbitrators (WIPO in particular) have extended their rulings into areas of law they have no right to enter
  8. Arbitrators have ignored legitimate ownership of domains
  9. The "bad faith" defence is a nonsense
  10. No one is monitoring UDRP

We'll cover the first five in greater depth here and the second five in a second part (save making this too long):

  1. It does not spell out who the burden of proof falls on
    In any civilised court, the decision lies with the accuser to prove that a person did what they say they did. Hence "reasonable doubt" in murder trials for example. This is a cornerstone in most law courts and as such is automatically presumed to be the case in arbitration decisions. It is, however, not explicitly stated and the last two years has seen a relaxation in what should be a hard and fast rule.

    [As several readers have rightly pointed out, this is for criminal courts only and civil courts work on a "balance of probabilities". Despite this, the point holds that UDRP does not provide an effective method for deciding guilt or innocence. Arbitrators are also guilty of ignoring the provisions that are in UDRP.]

  2. It does not say what evidence is required to prove innocence or guilt
    Again, a seemingly unimportant point but one of extraordinary import. Without a set approach to how the rules can be proven one way or the other, decisions made by judges are likely to vary hugely and so damage the creation of case law. It is also makes it far harder for a registrant to respond effectively to a challenge (see next point).
  3. It imposes impossible time limits
    The idea was that UDRP would be a fast mechanism for sorting out disputes. However, the 20-day reply period listed in the policy is ludicrously short. Worse than this, the 20-day countdown starts as soon as the arbitrator hears of the complaint - not the person that actually owns the domain.

    This means that if you aren't contacted by the complainant beforehand and then you don't check your email for a few days, you may only have a few days to: appoint a lawyer; write and submit your entire defence; and collect all the relevant evidence and documents you have (note no guidelines from above). On the other hand, the trademark owner will have had months to prepare.

    This puts all the pressure on the domain name owner. Also, in some hard-fought for cases, the arbitrator has refused further evidence from the defendant, despite it being extremely relevant, because the deadline had passed. Connected to this:

  4. A complainant does not need to contact a domain name owner before going to an arbitrator
    The first time someone may hear of a case is when the arbitrator email arrives in his inbox. Trademark holders have no need to contact a domain owner before going ahead. This is clearly daft and a court of law would take a dim view of someone going to court before attempting to sort out a dispute with someone in person - or even inform them of what they intended.

    Those that do contact domain name holders may also use it against the holder in a future case - there's a bit more of that in the point on bad faith (number nine).

  5. The complainant decides which arbitrator to go to
    This may seem fairly harmless since arbitration is mostly done over the Internet. However, due to the way the arbitration market has panned out, this simple choice has led to more inequality in domain dispute resolution than almost anything else.

    Arbitration has become big business. Thousands of cases are decided every year and money is taken for every one. Being the lead arbitrator for Internet domain names in a market of just four also provides a fair amount of power. Not to mention publicity when a high-profile case comes up.

    All of this has - and will continue to - give arbitrators a heavy incentive to rule in favour of complainants. The more decisions in favour, the more people will turn to you, the more money. This unchecked situation has resulted in some appalling decisions that have overriden not only UDRP but also international trademark laws.

    There was also recent evidence in which, due to a number of decisions made against AOL by WIPO, the massive Net conglomerate turned to alternative arbitrator NAF to rule on NAF promptly amazed all and sundry by deciding for AOL on the ridiculous pretext that the word "aim" at the start of the domain stood for "America Online Instant Messenger" - and that no one would think any differently.

    On top of this, arbitrators can decide themselves what judges they assign to what cases. There have been many calls for transparency in this system but so far nothing has happened. It is a fact that certain judges have been handed over 200 cases and have consistently ruled in favour of trademark holders. Other judges - on the books for an equal time - that have ruled against complainants have been on as few as ten cases.

Although not endorsed by him, special thanks goes to Michael Froomkin, who teaches at law in Miami and runs ICANNwatch, for his help and the supply of an as-yet unpublished document on UDRP.

Related Stories

Part II of the UDRP review is here
AOL wins rights to Aimster domain name

More about


Send us news

Other stories you might like