BT Response to US Government Green Paper:

A PROPOSAL TO IMPROVE TECHNICAL MANAGEMENT OF INTERNET NAMES AND ADDRESSES

[Docket No. 980212036-8036-01]

as published in the February 20, 1998 issue of the Federal Register (http://www.ntia.doc.gov/ntiahome/domainname/022098fedreg.txt)

About BT

BT is the UKís leading telecommunications network and service provider (see http://www.bt.com). BT has major international operations itself and offers competitive services in many countries through its numerous joint ventures, partnerships and alliances throughout the world. In Europe, particularly, BT is playing a major role in extending competition as telecommunications deregulation proceeds. BT is also at the forefront of the development and implementation of new interactive, on-line services and in the provision of the underlying infrastructure and technology to support such services. For example, through its Concert joint-venture (http://www.concert.com), BT is directly involved in upgrading the internet backbone to provide greater bandwidth and improved service.

BT offers internet-related services through initiatives including its BTnet business (http://www.bt.net), its consumer oriented BT Internet access service (http://www.btinternet.com), its website support service BT Webworld (http://www.btwebworld.com), the BT Campus range of educational products (http://www.campus.bt.com) and the general information and entertainment service Lineone (a joint venture business - see http://www.lineone.net).

In the course of its internet operations, BT is routinely involved in the registration and management of internet domain names, both on its own behalf and for its customers.

As a large, international and mainstream business, BT is also the owner of numerous trademark registrations for its products and services both in the UK and abroad. BT has been the target of a number of abusive misappropriations of its trade marks (including British Telecom and BT) in domain names registered by cybersquatters and others with no legitimate rights to use such marks. The ease with which names can be misappropriated in this way, and the threat of a disproportionate adverse impact because of the instantaneous global access provided by the internet, makes such abuses a serious concern for any commercial enterprise with a significant reputation and a market to protect. As a consequence, BT has had to threaten and to take legal action to protect its own rights on a number of occasions. In particular, BT was a driving force in the current leading UK case on cybersquatting (BT and others v. One In A Million Ltd. and Others - see http://www.nic.uk/judgment2.html).

BT has actively participated in discussions and commented on proposed changes to the systems of domain name administration, including the gTLD-MoU proposals (http://www.gtld-mou.org) and WIPO initiatives on trade mark and domain name issues, particularly regarding dispute resolution processes (http://www.wipo.int/eng/internet/domains/index.htm) as well as the previous NTIA request for comments (the previous BT response is available at http://www.wia.org/pub/bt-dnsnoi-comments.html). As a member of the International Trade Mark Association (INTA - http://www.inta.org) BT has also been closely involved in that organisationís deliberations on these matters.

However, to date, BT itself has not endorsed any of the existing proposals, such as the gTLD-MoU. Such proposals have included elements which BT considers would be worthy of support, but BT believes that as yet no proposal has been adequately directed at developing a framework for administration of domain names which will effectively meet the needs of future internet users, especially ordinary non-technical consumers and businesses, for a naming system which is equitable, predictable and stable.

Any changes should be made to promote user confidence in the internet as a medium for electronic commerce, not add to the already extensive potential for confusion and conflict over ownership and use of domain names.

Introduction

BT welcomes the US Administrationís participation and contribution to the debate on the future of the internet domain name system. The US administration clearly has a key role and responsibility in this matter and is in a position to strongly influence changes to improve the utility of the internet as a public resource for all.

And it is the wider mainstream constituency, the majority of which is yet to establish an internet presence and which is presently substantially unaware of the issues, which stands to be most affected by any changes. The impact on that constituency, therefore, demands the most careful consideration. The Administration should resist the clamour which appears to be coming mainly from technical members of the so-called "internet community" and those with vested interests in the sale of domain names (including Network Solutions Inc., certain gTLD-MoU CORE members and others who only see dollar signs in the Registry or Registrar businesses) for it to abdicate control and to make concessions primarily in their interest. (In this respect, it bears repeating that most claims to represent the "internet community" really only extend at most to those who are presently "internet aware").

The internet is at a critical inflection point in its development. It is in transition from adolescence to adulthood; from a stage where juvenile conduct could be tolerated because it seldom had serious consequences to a stage where more mature behaviour is demanded. The irresponsible indiscretions of adolescence cannot continue to be indulged with impunity. The potential for serious commercial repercussions is now too great. To use another analogy, it is no longer acceptable for the internet to be treated as an "electronic frontier" and administered subject to the crude principles of "frontier law" where the gunslinger with the fastest draw can impose terms regardless of the rights or interests of anyone else. A more civilised approach is needed.

The United States and other trading nations have well established laws to protect businesses and consumers from abuse. There are treaties establishing systems to facilitate fair trade and to ensure international respect for rights and freedoms. The system for administration of domain names likewise must provide sufficient order and safeguards to make the internet a civilised environment for international electronic trade. This will obviously require the implementation of appropriate rules by a responsible self-regulatory authority. The Administration is urged to establish practical requirements for such rules to adequately reflect existing real-world standards for commercial conduct.

More detailed comments on the proposals and issues raised in the Green Paper are provided below together with further explanations of BTís recommendations and conclusions. For convenient reference, BTís key recommendations and conclusions are summarised first.

 

Key Recommendations and Conclusions

Reform of Domain Name System Administration

 

The Need for Change

The Green Paper makes a number of assertions justifying the need for change. Whilst it can be accepted that most of the factors listed do contribute to the pressure for change, BT would question whether the priority accorded to those factors is accurate or entirely appropriate. Indeed, in our view, the Green Paper does not explicitly identify the single most critical reason for considering changes Ė that the existing domain name systems are patently not structured to meet the potential future demand which will come from the overwhelming majority of users who are yet to join the internet. Any changes ought primarily to be focussed on resolving this concern.

The Green Paper claims that there is widespread dissatisfaction with the alleged absence of competition in domain name registration. This statement is an oversimplification. There clearly is competition already between existing country code TLD registries and NSI as the gTLD registry. And that competition is not insignificant. As the RIPE CENTR comments record, ccTLD registrations comprise over a third of all domain name registrations (see http://www.ripe.net/centr/docs/response-green.html).

A more accurate statement of the competitive complaint would be that there is widespread dissatisfaction with NSIís for-profit, monopoly control over the existing gTLDs as registry and as sole registrar.

With respect to the other concerns listed in the Green Paper:

The future role of the US Government

It is clearly timely for the US Government to promote and oversee the transfer of responsibility for administration and operation of the domain name system to a suitably internationally constituted, independent body. The historical involvement of the US Government in the construction and administration of the internet places it in a unique position of authority to exert a stabilising influence on the evolution of its administrative systems.

Indeed, in BTís view, it is essential that the US Government should promote a smooth transition to an independent private-sector administration of a definitive domain name system which is adequately structured to meet future needs. However, the Administration should not relinquish its stewardship until it is confident that any changes to the domain name system will be properly directed at encouraging mainstream consumer and business confidence in the internet, will provide effective safeguards for existing rights and will not simply exacerbate the existing problems of conflicts and confusion.

 

PRINCIPLES FOR A NEW SYSTEM

  1. Stability
    As indicated above, BT fully endorses the view that changes should not be made "so quickly or depart so radically from existing structures" that the utility of the internet is adversely affected. Likewise, BT believes it would be undesirable to support competing root systems. Although it seems likely that the inherent flexibility of the internet will mean that certain technically competent groups will themselves implement alternative root systems (see e.g. the Grass Root Server proposals at
    http://www.best.com/~philh/grs), BT believes that such initiatives should be discouraged in view of the risk of conflict and fragmentation, particularly if different roots recognise different "registries" for the same TLD (cf. the competing claims for .web). Once again, for the benefit of consumer and business confidence, it is important that there should be one, stable, authoritative root system which can be trusted and which has accepted "official" status.
  2. Competition
    As also noted already in the comments on The Need for Change, BT would suggest that the relevance of competition in the various roles needs to be reconsidered. BT certainly supports the call for effective competition between Registrars (as distinguished from Registries in the Green Paper).

    However, we are sceptical of the proposal to introduce competing gTLD Registries. We doubt that this is either necessary or desirable and find the arguments against such a proposal more compelling. The ccTLD and gTLD Registries already do provide competition as well as complementing each other in some respects. In our view, it would be preferable firstly to promote changes in ccTLD administration to facilitate that competition (e.g. by encouraging ccTLD authorities to adopt consistent standards and frameworks for allocation of domain names). Despite assertions to the contrary from the advocates of adding gTLDs, the increasing number of ccTLD registrations demonstrates that many internet users still want to have a geographic indication in their domain names. Making it easier for them to do so would ease the pressure on the gTLDs.

    The competitive concerns over NSIís for-profit monopoly control of the existing gTLDs could be removed by reconstituting the gTLD Registry as a not-for-profit service (as is predominantly the case with the ccTLD registries). There are analogies to draw with other comparable "Registry" operations. Trade Mark Registries, for example, are operated as a public resource and no country operates competing Trade Mark registries in the same jurisdiction or suggests having one Registry to deal with trade marks in (say) Classes 1-10 and others to deal with trade marks in Classes, 11-20, 21-30, 31-42. Yet, if we consider trade mark Classes as corresponding to different gTLDs, this is effectively what the Green Paper seems to be proposing! There is no equivalent proposal for the introduction of more "registries" to compete with RIPE, ARIN and APNIC in their public service roles as regional IP Registries. Competition at that level is accepted as inappropriate. Competition is beneficial, and is rightly promoted between Internet Service Providers (ISPs) as access providers and as "Registrars" offering customers connection to internet addresses and registration of domain names.
  3. Private, Bottom-Up Co-ordination
    We agree that a private sector administration is likely to be more flexible and that it will be important to take account of input from users at all levels. However, as the internet matures into a mainstream medium, we take this to mean wider consideration for ordinary, non-technical business and consumer users and not just the internet technical specialists. The technical specialists have clearly contributed significantly to the development of the internet to date, and the value of that contribution is not to be underestimated, but now the internet is in transition to an international commercial medium, different priorities are appropriate. Just as, from the Wright brothers through to Boeing, many technical specialists have advanced the development of the aeroplane, but commercial air travel is not administered by or for the technicians, but for passengers and airlines.
  4. Representation
    As we have already stressed, the need for international representation, including all those who will be affected by proposed changes, must be recognised. This means that such representation must increasingly be drawn from mainstream commercial and consumer groups, not just the existing, internet-aware organisations with their predominantly technical and non-commercial bias.

THE PROPOSAL

The Co-ordinated Functions

The Administration proposes a new, not-for-profit corporation assuming authority over:

  1. policy for allocation of number blocks to regional number registries for the assignment of Internet addresses;
  2. the operation of an authoritative root server system;
  3. policy for determining the circumstances under which new top-level domains are added to the root system; and
  4. the development of other technical protocol parameters as needed to maintain universal connectivity on the Internet.

The allocation of number blocks and the development of protocol parameters appear to be principally technical issues, susceptible of resolution at a technical level. Clearly it is appropriate for RIPE, ARIN and APNIC and the ISPs who are their customers along with IAB to be concerned with the management of these functions. Provided this is done in a technically efficient and non-discriminatory manner which does not result in restrictions on access to or operation of the Internet, then it is unlikely to have a significant impact on the majority of businesses and consumers using the Internet.

In contrast, BT would endorse the proposition also presented by INTA, that the most important issues relating to administration of the DNS, including the policy for addition of generic top level domain names (gTLDs) and the guaranteed operation of an authoritative root server system are not primarily technical. Decisions on these functions will have a fundamental impact on businesses and consumers using the Internet. They will have significant consequences for the stability and interconnectivity of the Internet and particularly for its utility as an effective medium for electronic commerce.

In this respect, BT also believes that the Green paper has underestimated the damaging effects of loss of business and consumer confidence in the Internet as a result of changes which are likely to increase the potential for confusion and uncertainty over domain names.

In BTís submission, for reasons stated above and in its responses to previous consultation exercises, mainstream business and consumer representatives (including trademark interests) should logically have significant representation on any governing body responsible for domain name administration, and certainly no less than the technical community. The Internet technical community has consistently understated the relevance of trademark concerns to business and consumers. Trademarks are the lifeblood of international commerce and will inevitably play an increasingly important role on the Internet. As currently drafted, the Green Paper proposal does not provide for commercial interests to be adequately represented on the proposed Board of Directors, stating only that five seats on the Board "could be filled by commercial users, including trademark holders." This is not sufficient.

The Government or any "new corporation" will be disenfranchising the majority if it merely seeks the endorsement of the present "Internet Community" for any changes. Mainstream businesses and ordinary consumers must also have a proportionate say in discussion of those changes.

BT agrees with the view also expressed by INTA and other observers that management of domain name administration should be separated from management of the genuinely technical issues currently handled by IANA and others. Accordingly, BT also supports the INTA proposals regarding the composition of the Board of a "new corporation" insofar as it would be responsible for matters of domain name administration (and not just IP number allocation or protocol issues). Board representation should comprise the CEO of the new corporation together with, in proportion, two/three representatives of the registry/registrar community and at least seven members representing the international user community (present and future), including one seat for a non-profit organisation, one for individual users and the remaining five for mainstream business users, including trademark owners, from around the world. As use of the internet grows, we would anticipate that the new corporation would need to provide for a further proportionate expansion in its Board membership.

This proposed Board structure is consistent with the aim of ensuring Internet administration addresses the concerns of the businesses and consumers who will comprise the primary user community if the internet is to flourish as a medium for international commerce, education, entertainment and as a reliable source of information.

On decision taking, BT would agree with the Administration that decisions on key issues of policy, which would certainly include any decision to add new gTLDs, should require a super-majority or even complete consensus of Board members before implementation. It is important that critical changes should command as wide a support as possible.

The Competitive Functions

BT has serious misgivings about the Green Paperís proposals with respect to the introduction of competing Registries, its assessment of the trade mark issues and the suggestions for resolving the existing dilemma. The Administration does not seem to have appreciated the real nature and extent of the existing problems and to have pre-judged a number of key questions which ought to have been left for a fully representative "new corporation" to decide.

Our perception is that much of the pressure for change has in fact been driven, and continues to be driven, not primarily in the interest of developing a system which will realistically cope with future expansion and demand, but out of antipathy towards NSI in the internet technical community, supported by those with a direct financial interest simply in having more names to sell (e.g. many of the CORE registrars). They see their principal goal as breaking NSI's monopoly control over the existing gTLDs so that they can get a share of the profits they believe NSI must be making, regardless of whether they otherwise increase the risks of confusion and conflict over domain names.

Whilst removing NSI's registry monopoly may be a reasonable demand, proposals to achieve this by the creation of new registries and the simplistic, unstructured addition of new gTLDs fail to address the fundamental deficiencies in the Domain Name System and do not provide a logical path for orderly expansion to meet future needs. This is a key area where the gTLD-MoU plans fall down, as we see it, and the Green Paper proposals similarly attempt to sweep the issues under the carpet, in the naive expectation that "in the future, directory services of various sorts will make it easy for users to find the sites they need regardless of the number of top-level domains." Such effective directory services are not available yet. We do not need to add to the potential for confusion now just to encourage the development of directory services!

The Administration appears to have been swayed in NSIís interest to effectively endorse the proprietary registry model, allowing NSI to maintain its de-facto monopoly over the existing gTLDs (.com, .net, .org). NSI was originally given responsibility to manage these domains as a public resource. Many commentators have expressed concern over such monopoly control over a gTLD registry. BT shares such concerns and considers that the alternative, shared-registry model is to be preferred. In this respect, BT would endorse the pertinent principles expressed in the gTLD-MoU:

"(b) any administration, use and/or evolution of the Internet TLD space is a public policy issue and should be carried out in the interests and service of the public;
(c) related public policy needs to balance and represent the interests of the current and future stakeholders in the Internet name space
;"

Policy decisions ought not to be dictated by or carried out for the benefit of any one special interest group

BT endorses the arguments made by INTA in its submission on these matters.

The Administration's proposition that the proprietary registry model will encourage competition at that level is unconvincing. It will certainly expose the new corporation to competing claims and pressures to add more registries. Such pressures, particularly threats of litigation if new registries are not recognised, will almost inevitably result in a lowering of the entry criteria and ultimately in an unconstrained free-for-all. Trade mark owners will be faced with the practically insurmountable task of preventing abuse of their rights in countless gTLDs, subject to countless different registry dispute policies in countless different jurisdictions. The utility of the domain name system itself will be substantially degraded. Licensed anarchy will prevail.

The alternative single, shared registry model, with the registry operating on a cost-recovery basis and with competition between registrars represents a more stable, predictable proposition. Under this model, the new corporation would be empowered to manage the gTLD name-space as a public resource, would be able to take its decisions in the wider public interest and would not be involved in the dubious practice of granting proprietary monopolies to commercial enterprises to control individual gTLDs. The single registry model is also much to be preferred for dealing with disputes. Instead of having to deal with different dispute policies in different jurisdictions, businesses and challenged domain name holders would have a single consistent policy and jurisdiction would be clearly determined.

The Administration asserts that attempts to impose central order "risk stifling a medium like the Internet that is decentralised by nature and thrives on freedom and innovation". What the Administration appears to be reluctant to accept is that the inevitable consequence of unconstrained "freedom" (as apparently sought by the Internet technical community) is chaos, confusion and anarchy where only the street-wise (in this case hackers and others with sufficient technical skills) can survive. Already, the exploitation of this "freedom" by those without principles or any sense of community responsibility has given rise to the problems of proliferation of illegal content (e.g. child pornography), security breaches (e.g. hacking and spoofing incidents like Kashpureff's hijack of the root-server), spam and copyright infringements, all of which undermine established real-world standards.

The suggestion that introducing an orderly, formally structured and regulated system for domain name administration would somehow limit freedom of speech and expression is entirely spurious. The existing trade mark systems in the US and elsewhere do not abridge freedom of speech or restrict freedom in general simply because certain rules are applied to the registration of trade marks. It is a nonsense to suggest that imposing rules on the registration of domain names (which give a de facto exclusive right to each domain name) would do so.

Responsible authorities internationally, including in particular the US Administration must surely recognise that defence of "freedom" on the internet does not mean abdicating responsibility for fair and appropriate regulation. "Freedom" is not a licence to abuse the rights of others. In a civilised environment "freedom" must be exercised with responsibility. There is no good reason why a non-governmental self-regulatory body such as the proposed "new corporation" should not be able to operate sensible rules to manage the evolution of the domain name system as a public resource, and many good reasons why it should be encouraged to do so.

 

Creation of New gTLDs/ The Trademark Dilemma

The pressure to add new gTLDs has also come predominantly from members of the Internet technical community. The concerns of ordinary businesses and consumers have not been given weight commensurate with the potential impact the addition of new gTLDs might have on their interests.

The proliferation of further undifferentiated gTLDs will act to undermine the existing trademark system. We have already heard calls from members of the internet technical community for domain names to be above trademark law, yet for the value of domain names as on-line "brands" to be protected at the same time. Those making such calls evidently fail to appreciate the fundamental contradiction in that position.

BT strongly supports the view expressed by INTA on behalf of trade mark owners generally that it is undesirable and premature that the Administration should have already proposed to add up to five new gTLDs before the new corporation assumes responsibility. It is regrettable that the Administration has chosen to pre-empt the new corporation in this way and by stating explicitly that it (the Administration) stops short of not allowing any new top-level domains. BT believes it would be preferable to leave such decisions to the new, representative corporation. The unstructured addition of gTLDs will inevitably increase confusion and add to the risks for trade mark owners. New gTLDs should only be added, if at all, one-at-a-time and subject to appropriate safeguards, in the form of consistent, speedy and effective processes for administrative challenge and dispute resolution to protect existing trade mark rights.

Calls for new gTLDs are invariably based on the assertion that there are only a limited number of short, catchy names available in any given gTLD and that once these are taken, a new gTLD is needed to "give someone else a chance" to get the catchy names. The flaw in this logic is that as the number of essentially undifferentiated gTLDs grows, the assumed benefit of having a catchy name diminishes and the utility of the domain name system is substantially reduced. Just considering the existing and proposed gTLDs, for example, it is easy to accept that having bt.com is an advantage for a legitimate user of the name "BT". But what happens when new gTLDs like .firm, .web, or .shop come along? These gTLDs add nothing distinctive. Undifferentiated gTLDs can only add to what has been judicially recognised as the "inherent propensity to confuse". A new user would have no clues to distinguish bt.com from bt.web, bt.firm or bt.shop. In this situation, the simplistic First-Come, First-Served (FCFS) registration system (which in reality is First-Come, ONLY-Served) only compounds the potential for problems. The lack of distinguishing structure in the gTLDs and the FCFS allocation system mean that, if they can, businesses will try to register the same names in each new gTLD in order to strengthen their association with those names, to avoid dilution of that association and to exclude others as far as possible.

A more equitable and formal structure for the allocation of names and a move away from the inappropriate FCFS model is necessary if the system is to scale to meet future needs. The current essential uniqueness of domain names coupled with a simplistic FCFS system only encourages a "land rush" approach to grabbing of names to prevent others getting them or to hold potential users with legitimate rights to ransom for their preferred names. It is unsatisfactory that the first to register a domain name should be able to exclude any other business with legitimate real world rights from using the same name or mark in a domain name when it could not maintain such exclusivity in the real world. What is needed is a more sensible, orderly system which gives businesses and consumers confidence that the interests and rights of all (not just the fastest to register) will be properly considered and balanced. Domain names need to be as available for use and as readily distinguishable as trade marks are in the real world, whether a business comes early or late to the net.

BT would challenge the received wisdom propounded by those arguing for unstructured addition of gTLDs that potential domain name registrants would not accept a name linked to their specific goods or services. In the real world, businesses invariably use, and consumers are used to seeing, trade marks and names in context. Thus United Airlines and United Foods (say) would never be confused nor would they come into conflict in the real world, because they could not claim exclusive rights to the United name. But on the net, only one entity (and it doesnít happen to be either of those) can use united.com, and the domain name gives no assistance to allow the user to distinguish who is the owner until the relevant site has actually been visited or contacted. Confusion is inevitable if the relevant domain is not associated with the expected operator. For the inexperienced user who mistakes whitehouse.com for whitehouse.gov, a rude surprise awaits, for example.

In the real-world, consumers and businesses are used to systems (like Yellow Pages) which associate names with types of trade or services. There is no benefit to legitimate businesses who use the same mark for different goods and services if customers canít distinguish their websites (say MYMARK.firm and MYMARK.web) and waste time visiting both to find the right one. In circumstances where MYMARK is relatively common, then distinguishing by trade is the logical option.

BT would therefore advocate that new gTLDs should be added only where these distinguish classes of business and assist user access and where registration can be subject to objective tests for qualification (e.g. .airline might be established for IATA airlines with qualification determined by IATA). There is also a strong case for having a content-specific gTLD (and corresponding SLDs under ccTLDs) such as .xxx or .sex. Sexually explicit services could then be legally required to operate with domain names in this gTLD (or SLD under a ccTLD) which would make it much simpler and easier to control access to such sites to protect children from the whitehouse.com problem, for example. This would not be to impose censorship or restrict free speech, but would restore an effective means for consumer choice which sites like whitehouse.com subvert by exploiting that present "inherent propensity to confuse"). Similarly, non-commercial sites, including sites for abuse or complaint might usefully be allocated a specific SLD for that purpose.

The demand for more gTLDs to enable different entities to use the same basic name could also be reduced by having a shared (g)TLD which could itself be used as a directory. The Internet One initiative (see http://www.io.io) is an example of the sort of service which might serve such a purpose.

In BTís submission, new gTLDs should NOT be introduced unless they are structured and operated to overcome the shortcomings of the existing gTLDs. There is no sense in simply adding undifferentiated gTLDs (e.g. .web, .firm etc) because this will only replicate and multiply the existing problems. BT believes the Green Paperís proposal for the introduction of up to five new gTLDs of indeterminate nature, run by five separate registries will not resolve existing problems Ėthe competing Registries will have no incentives to operate structured gTLDs for the public good, rather they will only be motivated to maximise the number of names they can sell regardless of the longer term consequences for the efficient operation of the internet.

As BT has consistently argued, a more radical approach to reform is required. To this end, BT would propose that if any new gTLDs are to be created at this stage, they should be designed to test alternative models for dealing with existing concerns. This sort of testing is obviously more easily done by a non-profit registry, operating for the public benefit rather than one driven exclusively by the need to make a profit. Thus a new gTLD could be established for commercial use with SLDs for specific trade categories and with registration taking place at the third level (e.g. bt.telco.web). Likewise, another gTLD could be operated as a pseudo-directory with non-exclusive domain names (e.g. bt.dir might bring up a list of links to legitimate users of "BT" as a trade mark, or to a selection of different sites operated by this company.

It would also be possible to test the similar systems under ccTLDs. Indeed, it might first be preferable to do so Ė ccTLDs also offer the benefit that the trade specific SLDs under the ccTLD can be in the home language, respecting cultural diversity, which gTLDs are unlikely to be able to do so readily. Thus, for example, instead of the traditional .co.uk, .net.uk and .org.uk, it would be possible to operate a shared .dir.uk, and a UK specific selection of trade categories as SLDs (e.g. telco.uk). Presently, there is reportedly resistance to considering such proposals from those involved in domain name reselling who effectively control UK registry policy, based on the untested assertion that businesses will not want to use such specific categories. We do not accept that assertion. The present situation only allows one legitimate user the opportunity to register lloyds.co.uk, for example, although there is a bank, a chemist and an insurance business which all use the "lloyds" name. If separate business specific SLDs were available, each could have a domain name using lloyds without risk of confusion (in contrast, it is not immediately obvious who might operate lloyds.co.uk). To encourage ccTLD registries to adopt more business-oriented policies, we believe it would be desirable for ccTLD registries to consider establishing wider representation (as we have proposed for the Board of the new corporation), and including mainstream business and consumer groups in any policy setting body, which should be independent of the registry and registrars themselves.

The US Administration's proposal treats mainstream business, trademark and consumer interests as secondary, when it states "trademark/ domain name disputes arise very rarely on the Internet today." This statement is evidently based only on the experience of Network Solutions, Inc. (NSI) in litigation. Contrary to the Green Paper's assertion, BTís own experience and contacts suggest this is a significantly greater problem for businesses coming to the internet. There continue to be more and more cases every week. But reported cases represent merely the very tip of the iceberg. Anyone who discusses the issues with trade mark practitioners will be aware that there are many more instances where disputes have not gone to litigation, are unresolved, or have settled at significant cost to one, or both parties. Small businesses, finding their names already taken, frequently do not have the resources to mount a challenge and so reluctantly live with the situation and take another name.

BT agrees with the statement in the Administration's proposal that "where a trademark is unlawfully used as a domain name, consumers may be misled about the source of the product or service offered on the Internet, and trademark owners may not be able to protect their rights without expensive litigation." However, as appropriate as this statement is, we do not believe that it can be fully appreciated or addressed appropriately whilst there is an apparent unwillingness in the authors of the Green Paper to recognise that the underlying problems exist on a massive scale and are getting increasingly beyond control.

Catapulting the interests of domain holders over those of existing trademark owners can do little to increase business or consumer confidence in the internet as a vehicle for electronic commerce. Rather, it creates an inequitable and expensive structure in which tremendous goodwill and consumer confidence created in the "physical world" through substantial investment can be wrongfully exploited in the electronic world in the name of domain name holder "freedom" and "brand equity"! A first step towards resolving this situation is to respect trademark rights and the guarantees they provide for consumers

As any business which has ever sought to clear a trade mark for international use in the real world will be well aware, it is a time-consuming and expensive task. Invariably there are potentially conflicting rights in one jurisdiction or another, co-existence agreements to be negotiated and oppositions to be fought. Frequently, businesses spend very substantial sums promoting the marks they use and establishing an international reputation with associated recognition and goodwill.

Yet, in contrast, a business setting up on the internet can obtain a domain name with no check for existing rights and at little or no cost. With this domain name the business then gets instant access world-wide. If a business takes a domain name in an undifferentiated gTLD, some domain name advocates would argue that the absence of a country code affiliation or business identity in the gTLD, should enable the internet business to operate without challenge from a trade mark owner even if its mark is used in the domain name. That is surely an unreasonable expectation, given the immediate international reach of the internet, precisely because the relevant jurisdiction and trade cannot be identified from the gTLD. It cannot be acceptable that a business can claim an international presence on the internet and expect to operate with impunity even if it effectively usurps or trespasses on existing trade mark rights, when it could never do so in the real world.

Although NSIís dispute policy can certainly be criticised, it is surely correct in admitting challenge from trademark owners in any jurisdiction and in any trade. It is a recipe for conflict to claim the internet is a global medium and then argue that a domain name registrant should not have any responsibility to consider existing third party rights when using a name in a gTLD which gives no indication of the nature of a specific business or of any limitation on jurisdiction.

Administrative Issues

This inherent potential for conflict highlights the need for effective administrative procedures requiring the provision of sufficient information by domain name registrants and the operation of publicly searchable databases of registration details to enable trade mark owners to protect their rights against dilution and misappropriation.

The Green Paper acknowledges the desirability of taking appropriate steps in this direction. However, BT shares the view expressed by INTA that the Green Paper proposals fall short of what is really needed. BT therefore broadly supports the principles listed in INTAís statement regarding the minimum standards for application and renewal (see. http://www.inta.org/rspchk.tm). However, BT would suggest that updates should be required daily, and that any initial challenge period (e.g. the proposed 30 days) during which suspension is available should only commence from the date information on each new domain is actually posted to the public database.

Application and Renewal

1. The application must include:

2. Renewal

3. Other requirements

4. Publication

BT also strongly supports the INTA call for a single, consistent and effective administrative challenge mechanism for dealing with disputes as quickly as possible. A reasonable dispute resolution policy should be an essential element of any plan for the administration of domain names on the Internet. Discretion for domain name policy should neither reside with the domain name registrars, nor with the registries. Any dispute policy must be consistent across the gTLD space. A lack of uniformity and specificity will only lead to forum shopping by would be cybersquatters to take advantage of the gTLD registry with the weakest policy. If the goal is to create a system which is fair and predictable, and a relief from the current chaos, there can only be a single uniform system. There is no division among trademark holders on this point. A global marketplace and community requires a single set of global rules.

After much consideration, BT also believes that any dispute policy must avoid trying to resolve complex, fact-intensive trademark infringement disputes which are more appropriately suited to judicial analysis and decision.

Key Points for Dispute Policy

As already stated, BT disagrees with the Administration's view that "trademark/domain name disputes arise very rarely on the Internet today". Such disputes arise every day both at NSI and in the court systems of countries throughout the world, especially the United States, and are making up an increasing proportion of all trademark disputes.

BT believes that a dispute resolution system, which draws on the best features so far proposed in the gTLD-MoU policy, should be developed and given an opportunity to work. The gTLD-MoU policy is not yet entirely satisfactory, but it is a long way along the road towards being workable. Consequently, BT would recommend that the Administration gives due consideration to the hard work of the WIPO, IAHC and POC and build on that work.

Proposed Study

BT very much welcomes the Green Paperís commitment to undertake a study of the effects of adding new gTLDs and related dispute resolution procedures.

However, BT believes that it would be more appropriate for the study to comprise two parts. Firstly, there should be a detailed analysis PRIOR to the implementation of ANY new gTLDs (the Administration should withdraw its plan for the pre-emptive addition of five new gTLDs before such a study can be carried out). The pre-implementation analysis should include recommendations for the structure and operation of any experimental gTLDs. Secondly, a post-implementation appraisal should be conducted as the Green Paper suggests. No further gTLDs should be considered unless the study demonstrates that adverse consequences presently experienced by trade mark owners can be avoided

The Internet is an evolving medium, one with a profound effect on commerce and the law. Any changes to the domain name system will probably require continuous study and input for many years to come from all who are part of or affected by the Internet revolution.

However, it will be important to give the initial study credibility and to ensure that they are undertaken by an appropriately independent body not beholden to any one interest group. In this respect, the bill introduced by Senator Leahy (S.1727) is to be commended in proposing independent funding of the proposed study. BT cannot comment on whether or not the National Research Council would be the most competent or appropriate body to conduct the study, but we welcome the positive moves to back up the Green Paperís proposal for a detailed independent study.

The Transition

BT reiterates that it is no satisfactory for the Administration to pre-empt the new corporation by committing it to introduce five new gTLDs and five new registries before it has even been formed. As previously stated, BT would urge the Administration instead to speed up formation of a business and consumer-oriented Board and to leave all decisions on changes to be decided by the new corporation in the light of the preliminary study also proposed above.

Competitive Registries, Registrars and the Addition of New gTLDs

As discussed above, BT favours the alternative single, shared registry model, as proposed under the gTLD-MoU plan. BT shares the view that consumers will benefit sufficiently from competition among market-oriented registrars.

BT understands the Administration's interest in introducing competition in the marketplace. However, we would repeat that the introduction of up to five new registries, with each having administrative responsibility for one new gTLD, before there are procedures for dealing with issues of misappropriation of trademarks, is premature. For the reasons discussed above, adequate safeguards must first be in place before additional gTLDs are introduced. Accordingly, BT urges the Administration to defer the introduction of any new gTLDs until the trademark issues have been adequately addressed.

The Root Server System

BT strongly supports proposals to introduce professional management and to upgrade the security of the root server system. It is essential to the stability and interconnectivity of the internet that there should be a single definitive root server system which is run according to properly documented procedures and operated by professionals who are fully accountable to the appropriate authority (i.e. the new corporation) for their actions. In no circumstances should it be possible for individuals to take unilateral decisions affecting the root server system. There must be no repeat of the Kashpureff hijacks or equally of the "test" reroutings recently conducted as a personal initiative by Jon Postel. Such incidents only serve to lower public confidence in the operation of the internet.

The .us Domain

BT encourages further study of expanded use of the .us domain space. The .us namespace is the only one of which it can truly be said that the .us domain name is equivalent to an ordinary address. BT shares the Administration's analysis that the "locality-based" system may not be attractive to commercial users. BT would agree that there would be advantages to opening up greater use of the .us domain, particularly along the lines we have already suggested for expanding ccTLD namespace, and assuming adequate safeguards are put in place for dealing with the misuse of trademarks.

[END]

Comments submitted 23 March 1998

Contact: Keith Gymer

BT Group Legal Services Intellectual Property Department

8th Floor Holborn Centre

120 Holborn

London EC1N 2TE

UK

Email comments to: keith.gymer@bt.com