Domain Names I Governance and Ownership

1
Domain Names (I) Governance and Ownership

• Internet Governance Topic 5
• Professor Graham Greenleaf and Russell Allen

2
Control of domain names

• Allocation
• In theory, a global hierarchical structure
• Crucial control is of Top Level Domains (TLDs) -
  both generic (gTLDs) and by country (ccTLDs)
• ICANN controls .com and .org gTLDs, and creation
  of new ones
• For ccTLDs, the control structure differs by
  country
• At every DNS level, someone has the right to
  allocate new sub-domains
• DNS servers located throughout the Internet, keep
  at least the addresses of the root servers for
  each TLD

3
Legal implications of domain names

• Uniqueness gives commercial value
• How TLDs are managed is very political - it
  controls who can be found on the net
• Eg Chinas attempt to control Chinese character
  names
• Eg attempts to create .sucks and .union TLDs
• Threats and attempts to create alternative DNS
  systems result
• Disputes within ICANN

4
Control of domain names (2)

• Most disputes are about domain name allocation
• We assume domain names will function
  internationally once allocated -
• that DNS servers will resolve to the correct
  IP address
• What type of rule or law is this?
• Exceptions
• Chinas corruption of the DNS to make
  ltgoogle.comgt resolve to Chinese search engines

5
Are Domain Names Property?

• Domain names are merely contractual arrangements.
• Maybe linked, however, to Trade Marks
• Network Solutions Inc v. Umbro Intl Inc. 529 SE
  2d Domain names cannot be garnished in
  bankruptcy
• Zurakov v. Register.com (Unrep, Supreme Court of
  State of New York, Maskowitz J, 27 July 2001)
  Merely contractual

6
Organisational Control

• http//www.artslaw.com.au
• ICANN (.), delegates to
• auDA (.au), delegates to
• MelbourneIT et al (.com.au), delegates to
• Arts Law Centre of Australia (artslaw.com.au),
  delegates to
• Itself! (www.artslaw.com.au)

7
Can we avoid the entire system?

• Alternate Roots are Technically Possible
• Alternate Roots have existed - .bix, .kids etc
• but, none has got critical mass to interfere
  with ICANN policies.

8
Allocation Pre-DNS

• Single list of names, shared between all
  computers.
• Didnt scale
• New System was needed.

9
Some Terms

• TLD Top Level Domain
• gTLDs - .com, .org etc
• ccTLDs - .au, .uk, .nu
• specialTLD - .arpa

10
gTLDs

• sTLD and uTLD
• In 1980s
• .org .com .edu .gov .int .mil .net
• Now
• .aero .biz .coop .info .museum .name
  .pro

11
ccTLDs

• Countries or merely indicators?
• Over 240 countries
• Changes over time

12
Allocation Pre-ICANN

• IANA, Jon Postel and NSI

13
The Creation of ICANN
14
ccTLD Issues

• Nuie Case Study
• .au Degation Changes Case Study

15
Issues

• Dispute resolution policies (DRPs)
• gTLD, .hk, .cn, .au
• Criticisms of DRPs and their administration
• Decisions under DRPs - law or chaos?
• Chinese (and other) character domain names
• ICANN as a model of Internet governance?

16
Domain Names (II) gTLD Dispute Resolution
17
gTLD Dispute resolution policies (ICANNs UDRP)

• Substantive grounds - P4(a), (b) and (c)
• Choice of law - no provision - R15(a)
• 1 or 3 member panel? -
• P4(d), R3(b)(iv), R5(b)(iv), P4(g) (fees)
• Complainant (TM owner) chooses if only 1
• Both complainant and respondent can request 3
• No appeal provisions (even if only 1 chosen)
• Remedies limited to cancel/transfer (P3, 4(i))

18
gTLD Dispute resolution policies (ICANNs UDRP)

• Court proceedings
• Reg will obey order of a Court/Tribunal of
  competent jurisdiction (P3(b))
• P4(k) - either party can submit to Court, either
  before or after Panel decision (or 10 day stay)
• R3(b)(xiii) - complainant agrees to submit to
  jurn of Court in Mutual jurisdiction (R1)

19
Criticisms of UDRP and its administration

• Dr Milton Mueller Rough Justice (2000)
• Suspected incentive for providers to favour
  complainants to attract cases (they choose)
• Proved substantial results bias of 2/3 providers
  in favour of TM owners
• Prof Michael Geist Fair.com? (2001)
• Shows how the results bias occurs
• Non-decisive factors
• Price competition is very limited
• Panelists of the 3 providers were very similar
• Advertising and supplemental rules also minor

20
Geists Fair.com? criticisms

• Decisive factor is case outcomes - but why are
  the outcomes biased?
• Allocation of panelists provides the answer
• Little known about allocation procedures
• However, study of 3,000 allocation outcomes show
  the 2 pro-TM providers
• Allocate cases only to a handful of panelists
• These result in 83 decisions pro-TM owner
• For 3 member panels, drops to 60

21
Suggested UDRP reforms

• Geists Fair.com? suggests
• Mandatory 3 member panels (complainants cost)
• Caseload maximum/minimum for all panelists
• Quality control reviews of panelists
• Better transparency through searching
• Meullers Rough Justice suggest
• Random selection of panelists
• Appellate procedure
• Closer ties between registrars and providers

22
UDRP caselaw - Overview

• How do you find it? - implications
• Elements of a UDRP complaint
• DN identical or confusingly similar
• Registrant has no rights or legitimate interests
• Evidence of bad faith
• Decisions under .cc DRPs
• DRP decisions challenged in Court
• Other Court decisions on domain names
• Are the UDRP and .ccDRPs successful?

23
Finding decisions - the problem

• Number of decisions to consider
• Already 1,300 gTLD decisions in 2002
• Usually no system of appeals - single instance
  wilderness
• Number of decision-makers
• gTLDs have 4
• Most .ccTLDs have one or more
• Courts add to the numbers
• Panelists must (and do) consider them
• No formal system of precent, just good practice
  and efficient to consider expertise of others
• Unusual for arbitrators / mediator to have so
  many decisions by their peers to consider
• Result is that secondary analyses are important

24
Existing search facilities

• Berkman Centers UDRP Opinions - search engine
  with limited connectors etc
• ICANN UDRP search - uses Berkman
• Berkman Centers UDRPmate - hypertext linking
  between decisions by cutnpaste
• Umass UDRP seach engine - complex searches
  possible
• UMass/Cornell LII UDRP Publishing Project -
  searching based on detailed subject indexing
• See lthttp//lweb.law.harvard.edu/udrp/library.html
  gt

25
Studies of UDRP decisions

• Early analysis by Cabell (Harvard) 2000
• Kur UDRP, 2002, Max Planck Institute
• analysis of over 3,000 decisions to mid-2001
• No major flaws in principles applied (but
  considers eResolution discrepancy needs
  addressing)
• Main problem is confusingly similar cases
  (free speech implications issues of proof) -
  does/should UDRP require actual confusion?
• Possible procedural reforms defaults and appeals
• Mueller and Geist - more statistics than law

26
Limited role of UDRP decisions, and burden of
proof

• UDRP limited to disputes with 3 elements
• Disputes outside these elements must be resolved
  by Courts, not UDRP panels (deference by UDRP).
  This is crucial to proper interpretation.
• cl 4 - burden on complainant to prove all 3.
• But cl 4(c) seems to require respondent to
  establish legitimate interest
• Kur Complainant must establish prima facie case,
  then burden shifts to respondent

27
Elements of a complaint

• Complainant must prove all 3 (cl 4(a))
• (i) domain name is identical or confusingly
  similar to complainants TM
• (ii) registrant has no rights or legitimate
  interests in respect of the domain name
• (iii) domain name has been registered and is
  being used in bad faith.

28
DN identical or confusingly similar to TM

• UDRP 4(a)(i)
• Few cases - usually identical
• The sucks cases
• Lockheed Martin v Paresi 2000
• Vivendi v Sallen 2001
• Two views of the UDRP? (see Kur 3.2.1)
• (I) actual confusion not necessary
  (misappropriation), but lack of bad faith may
  excuse (separate ground)
• (ii) actual confusion is necessary - leave
  other variations to the Courts to decide
  (deference)

29
DN identical or confusingly similar to TM

• Mass registration cases
• Eg 38 variations of ltyahoo!gt all found
  confusingly similar (Kur)
• Fan sites and sales sites may be confusingly
  similar but not in bad faith

30
Bad faith - Registration and use

• UDRP requires DN 'has been registered and is
  being used in bad faith' (cl 4(a)(iii))
• Requirement of use has been weakened
• Offer to sell considered to be use
• Passive holding (inactive use) included
• Telstra Case

31
Evidence of bad faith

• UDRP 4(a(iii) and 4(b)(i), (ii), (iii) or (iv)
• Importance of prior knowledge of TM
• (i) Registration for gain from TM
  owner/competitor
• (ii) Pattern of hoarding names
• (iii) Intent to disrupt business of competitor
• (iv) Intent to divert users to website by
  confusion
• Other evidence of bad faith

32
Prior knowledge of TM

• The most essential aspect is not even
  mentioned expressly in the Policy - whether
  respondent was aware or had reason to be aware of
  conflicting TM (Kur 3.2.2.1)
• Usually prima facie sufficient for bad faith
• Respondent then has 2 options
• Refute claim of knowledge by showing DN chosen
  for credible independent reasons (no bad faith)
• Show positive prior legitimate interest in the
  DN, so issue of knowledge becomes irrelevant

33
Bad faith - (I) intent to sell to TM owner /
competitor

• Registration with intent to profit from TM
  owner/competitor (cl 4(b)(i))
• Intent to sell to others irrelevant
• Must be intent to sell to a complainant with a
  reputation in the name

34
Bad faith - (II) Pattern of conduct of hoarding
names

• Registering name to prevent TM owner, if part of
  a pattern of hoarding names (cl 4(b)(ii))
• Only names relevant are those in which there is a
  prior reputation
• Avoids need to show intent to sell, if hoarding
  can be shown

35
Bad faith - (III) Intent to disrupt competitors
business

• Registration with intent to disrupt competitors
  business (cl 4 b(iii)
• Examples
• Diversion of DN to an undesirable site
• Diversion of DN to any other website to cause
  confusion
• Typo diversions - obtaining a (similar) DN to
  attract typo error users to your site or another

36
Bad faith - (iv) Diversion to your own site, by
creating confusion

• Using DN to intentionally divert users to your
  site for commercial gain, by creating confusion
  with complainants TM (cl 4(b)(iv))
• Eg Reed Executive v Reed Business Information (UK
  HC, 2002) - use of metatag Reed (same as TM of
  other party) was a breach of TM - implications
  for UDRP, as metatag could increase confusion
  caused by a similar domain name

37
Bad faith - (I) Other indicators

• UDRP cl 4 - factors not exhaustive
• Panels have inferred bad faith from failure to
  respond and little else

38
Registrant has no rights or legitimate interests

• Default advantage of 1st registration
• Holder should win if has legitimate interest
• UDRP 4(a)(ii) and 4(c)(i), (ii) or (iii)
• Use re bona fide goods/services (I)
• Distributors, licencees and resellers
• Use of own name (ii)
• Honest concurrent use (iii)

39
Legitimate use - for marketing

• Sale/advertising of domain name itself can
  satisfy this (if before notice of TM)
• Distributors, licencees and resellers
• have legitimate uses of TMs to advertise the
  goods/services provided
• They have notice of mark, not of dispute
• bona fide offering must not extend to
  pretending to be the TM owner
• Protection against similar names only
• Descriptive /geographical/ generic names
• More likely to succeed the more generic is the TM

40
Legitimate use - Honest concurrent use

• First to register DN of honest concurrent users
  should be able to retain name
• Any further disputes must be left to Courts
• Examples
• Fuji Film v Fuji Publishing -

41
Legitimate use - Own name

• Examples
• Penguin v Katz - nickname Penguin was enough to
  retain DN
• Change of name to Mr Oxford University was not
  enough to retain DN

42
Legitimate use - other

• Cl 4 c (iii) allows legitimate noncommercial or
  fair use, not to mislead consumers or tarnish
  the TM

43
UDRP decisions in Court

• Sources of decisions
• WIPO Selection of UDRP-related Court Cases - 16
  to Oct 02
• Jones UDRP "Appeals" in Court (on UDRP.net) -
  61 to Oct 02
• Types of analysis
• Patrick Gunning - do Courts decide differently
  from UDRP panels?
• David Sorkin - what deference do and should US
  courts show to UDRP panels?

44
UDRP - Do Courts decide differently?

• Decisions same as UDRP panel
• Strick Corporation v Strickland (US Dist Ct)
• Victorias Cyber Secret (US Dist Ct)
• Barcelona.com Inc (US Dist Ct)
• Decisions contrary to UDRP panel
• Storey v Cello (US Dist Ct) - procedural issue
• Gunnings conclusion - too early to say
• Few decisions appealed, and Court usually
  reached same conclusion as UDRP panel
• But this could also reflect cost of litigation
  and .com downturn

45
UDRP - What deference should Courts give it?

• Sorkin US Courts normally apply an extremely
  deferential standard of review to arbitration
  decisions
• Reviews 5 UDRP decisions before US federal courts
  - UDRP decisions ignored or briefly mentioned
  some Court said that decision would be de novo

46
UDRP - What deference should Courts give it?

• Sorkin concludes review should be de novo
• UDRP excludes legitimate disputes
• (anational) UDRPolicy ? trademark law
• Procedural reasons are even stronger
• Parties expectations (and intent of UDRP)
• Streamlined UDRP procedures do not allow testing
  of evidence
• Disparity in appeal requirements deny due process
• Conclusion de novo review is consistent with
  UDRP and required for fairness

47
Domain Names (III) ccTLD policies on Dispute
Resolution
48
.hk Dispute resolution policy

• HKDNDR appoints HKIAC sole provider
• Same as gTLD UDRP except
• Only orders of a HK Court are recognised (P3(b))
  - enforcing a foreign judgment?
• Similarity must be to a TM in HK (P4(a)(i))
• R3(b)(xi) requires complainant consent to final
  and binding arbitration see also P4(a)
• Arbitration Ordinance s2AC requires written
  agreement
• Same provisions for 1 or 3 member panels

49
.cn Dispute resolution policy

• Covers both .cn and CNNICs Chinese character
  names
• CNNIC appoints CITEC as service provider
• Similar to gTLD UDRP in most respects
• Substantive grounds similar (P5, P8, P9)
• Can submit to a Chinese court or arbitration body
  at any time (P14, P15)

50
(No Transcript)