What is the Legal Status of Hyperlinking and/or Deep Linking?

Michael Travis BA/LLB (Hons) 2002, Auckland University Law Review Please notify the author if you wish to cite this paper, or use extracts in your work. A PDF is also available via Google Drive.   “The deep linking issue attempts to answer the question that’s been asked since the Internet first became part of the general public’s consciousness: Is this medium a free source of information for the benefit of the people, or a controlled presentation of branded content that benefits commercial interests?”[1] Hyperlinks       On the internet, every public webpage is available to the world through a Uniform Resource Locator (“URL”).  A URL constitutes the address of a webpage and allows that page to be loaded into a web browser.  Because a URL defines the location of a particular webpage, other pages are able to list that URL in the form of a link.  By following the link, the web browser will change address and so load a new webpage.[2]  This is known as hyperlinking.  Hyperlinks are commonly represented by underlined text, but can take any form, including graphics.  Links can be made to websites, individual web pages or files within a website, and even to different positions within a single webpage (for example a link at the bottom of a page to return to the top of the page). 1.      Deep Linking       Websites have a particular structure.  There is an initial directory containing index files, which form the content of the website’s front page (or homepage).  There may then be any number of subdirectories containing further information to be displayed as a person browses deeper into the website.  Since hyperlinks can point directly to any particular file or page within a website’s structure, one can easily bypass a homepage and direct a web browser deep into the contents of a website.  A hyperlink which links directly to such an item, instead of the homepage or initial directory, is known as a deep link.  In recent times, deep linking has become the subject of great controversy.  Certain internet users and website owners have begun to question the legality of hyperlinking in general, and deep linking in particular. Controversy 1.  Objections to Deep Linking There are various reasons for which website owners might object to the use of deep links by others to link directly to specific pages or files within their websites.  For example: Some owners derive revenue from advertising placed on the front pages of their websites, and consequently suffer loss when those pages are circumvented; Some owners provide content which can only be accessed after certain information has been submitted, for the purposes of: (a)    security; (b)   marketing; (c)     contracting; or (d)    other legal purposes; and these objectives are compromised when such systems are circumvented; Some links may be presented in such a way as to offend the intellectual property rights of website owners or others; and Some owners simply “expect to be able to control every facet of how … users access their services and content”.[3] Similarly, Charles Conn, CEO of TicketMaster, explains that his company “is in favour of linking” but disapproves of “competitors linking to all of a site’s content to build a business on the back of [another] company’s site”.[4] In some instances, website owners have complained that they suffer adverse effects as a result of being targeted by linkers.  For example, eBay filed an application for an injunction against Bidder’s Edge,[5] which was using spider software[6] to regularly search eBay’s online auction website and was then processing the information to form deep links.  Judge Whyte granted the injunction, holding that the searches by Bidder’s Edge “diminished the quality or value of eBay’s computer systems” by “consum[ing] at least a portion of [the] plaintiff’s bandwidth and server capacity”, thus slowing, or having the potential to slow, eBay’s site.[7] This decision was cited favourably by the Court in Ticketmaster Corp v Tickets.com Inc.[8]  In that case, however, Judge Hupp found that the use of spider software by Tickets.com was “very small” and that there was “no [evidence] showing that the use interfere[d] to any extent with the regular business of [Ticketmaster]”.[9]  Nonetheless, “[i]f it did, an injunction might well issue”.[10] 2.  Support for Hyperlinks The very nature of the internet is that it is a series of interconnections between computers and networks.  Linking supporters argue that any restrictions on this process would destroy what makes the internet a unique system: “[I]f you stop people from linking then the Web is no longer a Web.  It would become a collection of isolated chunks of information.”[11] Moreover, the internet has evolved as a public medium where linking is a recognised and commonplace process.  Objectors to deep linking “want their pages to be openly available to individual visitors but not to other sites – a division rendered nearly impossible by the very technical structure of the Web”.[12]  And if a webpage is available to the public, should it not be equally available to other websites? Tim Berners-Lee, the creator of the internet, explains that: “The ability to refer to a document (or a person or anything else) is in general a fundamental right of free speech to the same extent that speech is free.”[13]  Deep linking benefits the internet user because “[m]aking the reference with a hypertext link is more efficient but changes nothing else”.[14] In most cases, linking can also benefit the referred website.  As the flow of web browser traffic to the referred site increases, so does its hit record.[15]  This encourages increased custom, publicity and advertising revenue as well as superior positioning on many search engines.  When faced with a lawsuit in the Ticketmaster case, W. Thomas Gimple, President and CEO of Tickets.com, explained that their use of deep linking was “a service to consumers and venues, and brings customers to Ticketmaster.  So we’re a little puzzled about this.”[16] 3.  Balance of Convenience In any event, it is possible for website owners to structure their content to prevent, or otherwise hamper, others from deep linking to their web pages.  Laszlo Pataki, a website designer, asks, “[w]hy bring the lawyers in when there are simple technological fixes…?”[17] For example, websites using frameset commands[18] prevent web browsers from displaying the locations of individual web pages in the address field.  Instead the browser displays a static URL (usually the address of the initial directory, or homepage).  With URLs thus masked, links are harder to make.  Other types of website programming conceal the locations of individual pages and files.[19] One alternative development is the Robot Exclusion Protocol, a system that prevents the making of undesired links by automated programs.  Jakob Nielsen writes that, by simply including a specific meta-tag in a webpage’s programming, “[w]ell-behaved search engines will exclude any such page from their databases”.[20]  Online auctioneer eBay uses such ‘robot exclusion headers’.[21]  Similarly, there have been proposals for the establishment of a ‘deep linking exclusion protocol’, which “says whether or not links are permitted into the website, and further, if linking is permitted … which sub-directories may be linked to”.[22] 4.  Case Law There have been many disputes dealing with various aspects of the hyperlinking debate.  Unfortunately, though many are litigated, the majority settle out of court and there is little established in the way of useful case law.  In those few instances where judgments have been issued, all deal with specific aspects of the debate[23] and none have clarified the legality of hyperlinking, or deep linking, in and of itself.  As a result, it is presently impossible to predict the outcome of a test case.  Certainly, academia has been vocal in support of existing internet protocols[24] but it is altogether likely that courts will eventually limit the practice of hyperlinking to some degree. That said, the general tenor of cases seems to suggest that hyperlinking itself is not the problem, but rather the motivations behind it, or the exact manner in which it is done.  A most instructive decision in illustrating this point is again one of Judge Hupp’s in Ticketmaster,[25] where the Judge’s comments in rejecting several grounds of complaint prove insightful. Firstly, Judge Hupp dismissed a claim of copyright infringement on the grounds that:[26] [H]yperlinking does not itself involve a violation of the [American] Copyright Act … since no copying is involved, the customer is automatically transferred to the particular genuine web page of the original author.  There is no deception in what is happening.  This is analogous to using a library’s card index to get reference to particular items, albeit faster and more efficiently. Secondly, the Judge dismissed a claim of breach of contract.  While Ticketmaster displayed terms and conditions for deep linking on its website, there was no proof that Tickets.com knew of or agreed to these conditions.  Thirdly, the Judge rejected an argument that deep linking was an example of unfair competition, concluding that “deep linking by itself (i.e., without confusion of source) does not necessarily involve unfair competition”.[27] With this case and others in mind, the following recommendations describe a practical approach to hyperlinking that takes into account both accepted practice and precautionary legal considerations. Recommendations for Hyperlinking As a matter of policy (if not law) hyperlinking should be avoided outright where it: 1.       Adversely affects the linked site; 2.       Constitutes database infringement (prohibited in Europe); or 3.       Amounts to a breach of contract. A claim of breach of contract presupposes the existence of a contract between the linked party and the linker, prohibiting the making of links.  How might such a contract be formed? In most instances, links are made by visiting the target web page and copying the address field.  In such a situation, successful contract formation may depend on some form of interactive access to the site, whereby assent to terms and conditions is evidenced by a positive act.[28] By contrast, merely displaying the terms and conditions on the website is probably insufficient to bind users,[29] as is ‘security by obscurity’ (that is, the assumption that it is illegal to link to a well hidden address by virtue of its secrecy). 1.  Referencing Tim Berners-Lee explains that: “The intention in the design of the web was that normal links should simply be references, with no implied meaning.”[30]  This reflects the opinion of the World Wide Web Consortium,[31] which writes that “[l]inks are merely references to other sites.  You don’t have to ask permission to link to … any other website.”[32]  Just as there is no legal precedent to prevent referencing in the world at large, there is no reason in law why there should be a prohibition against hyperlinking per se.  Nonetheless, while Berners-Lee argues that there “is no reason to have to ask before making a link to another site”, we are responsible for what we say about other people and the content of their sites “on the web as anywhere”.[33] There are various established limitations on our freedom of speech to make references that must equally apply to hyperlinking.  Thus, for example, a hyperlink should not falsely or inaccurately imply: 1.       Endorsement; 2.       Common authorship; 3.        Confusion of source; or 4.       That one document is to be considered part of another. It is important that any person making a link clarifies the relative status of the linked document. 2. Processing a Hyperlink The general procedure for linking already safeguards against false suggestions of authorship and other matters.  Typically, following a link will either create a new window in which the linked document is displayed, or the linked document will be loaded in the same window so as to replace the original document.  Of these, the former is preferable as: “The window system has a user interface metaphor that things in different windows are different objects.”[34]  This helps clarify where the author of the original document is not the same as the author of the linked document (especially if the link is also appropriately labelled).[35] ‘Framing’ poses particular issues.  This process allows the web browser to display multiple web pages at the same time by dividing up the browser window into subsections (or frames).  Since following hyperlinks merely replaces part of the browser window with the linked document, “it is easy to create the impression that the owner of the surrounding frames is in fact responsible” for that document.[36] Despite the highly disparate nature of the case law, courts have shown themselves to be uniquely allied in prohibiting framing where the linker does not have the permission of the linked party.[37]  Consequently, this choice of interface should be avoided where permission has not been obtained. 3.  Describing a Hyperlink Similarly, one should be cautious as to how a hyperlink describes the linked content.  Incorrect or false labelling may create an action in: 1.       Tort (for example defamation); 2.       Passing off; or 3.       An aspect of intellectual property (for example, paternity or integrity). Conclusions The great hope of the devout internet community is that the legal system will eventually come to “understand and honour the Web’s essential openness, while leaving room for the law to crack down on truly parasitical behaviour”.[38]  Unfortunately, it seems that this balance lies somewhere in the distant future.  At present, there is no legal precedent establishing either an individual’s ‘right to link’ or the right of a website owner to prohibit links – yet more and more websites are introducing online linking policies forbidding linking or framing without prior written consent.[39]  Likewise, the inevitable pressure on lawmakers from commercial interests may result in determinations that privilege the online industry over end users. Nonetheless, the internet has proven itself to be adept at self-regulation.  In the absence of legal intervention, accepted linking practices will no doubt evolve and can thus provide a workable guide to the courts in the event of litigation.  [1] Darren Deutschman, a legal consultant on intellectual property rights, quoted in Finley, “Attention Editors: Deep Link Away” at Wired.com <http://www.wired.com/news/politics/0,1283,35306,00.html> (last accessed 6 September 2002). [2] Instructions to the web browser are stored in the hypertext code (HTML) of each webpage. [3] Rosenberg, “Don’t Link or I’ll Sue” <http://www.salon.com/tech/col/rose/1999/08/12/deep_links> (last accessed 6 September 2002). [4] Quoted in Finley, supra note 1.  Similarly, owners complain where competitors “piggyback on us under the guise of community service”: Alan Citron, President of Ticketmaster Multimedia, quoted in Macavinta, “Sidewalk Sidesteps Ticketmaster” at CNET News.com <http://news.com.com/2100-1023-279913.html> (last accessed 14 May 2002).  [5] eBay Inc v Bidder’s Edge Inc 100 F Supp 2d 1058 (ND Cal, 2000). [6] Automatic or robotic computer script or program “which operates across the Internet to perform searching, copying or retrieving functions on the web sites of others”: Ibid 1060. [7] Ibid 1071. [8] 2000 US Dist LEXIS *12987 (CD Cal). [9] Ibid 17. [10] Ibid. [11] Laszlo Pataki, a website designer, quoted in Finley, supra note 1. [12] Supra note 3. [13] Berners-Lee, “Links and Law: Myths” <http://www.w3.org/DesignIssues/LinkMyths.html> (last accessed 6 September 2002). [14] Ibid. [15] This records the number of times the website has been loaded remotely, which roughly corresponds to the number of new visitors to the site. [16] Quoted in “Ticketmaster Sues Again Over Links”, The New York Times on the Web, 10 August 1999    <http://www.nytimes.com/library/tech/99/08/cyber/articles/10tickets.html> (last accessed 14 May 2002). [17] Quoted in Finley, supra note 1. [18] A hypertext command that divides the browser window into multiple frames.  This process is described in further detail below. [19] Java <http://java.sun.com>, Flash <http://www.macromedia.com> and Perl <http://www.perl.com> (last accessed 6 September 2002). [20] Nielsen, “Deep Linking is Good Linking” <http://www.useit.com/alertbox/20020303.html> (last accessed 6 September 2002). [21] See eBay’s User Agreement 19 April 2002, para 7 at <http://pages.ebay.com/help/community/png-user.html> (last accessed 6 September 2002). [22] Winer, “Deep Linking” <http://davenet.userland.com/1999/08/09/deepLinking> (last accessed 6 September 2002). [23] For example, the use of ‘framing’ in linking: Infra note 37. [24] A brief was filed by 28 Professors of Law (as amici curiae) supporting reversal of the decision in eBay Inc.  They complained that “the court’s rationale sweeps so broadly as to endanger many of the most fundamental activities on which the Internet and electronic commerce are based” (p 2).  Available at <http://www.law.berkeley.edu/institutes/bclt/pubs/lemley/bedgeami.pdf> (last accessed 6 September 2002). [25] 2000 US Dist LEXIS *4553 (CD Cal). [26] Ibid 6. [27] Ibid 9. [28] For example, clicking an ‘acceptance’ button or icon.  This is known as a click-wrap agreement (also web-wrap or browse-wrap).  The binding legal force of such agreements is itself the subject of much debate. [29] See Ticketmaster, supra note 25, 7-8. [30] Berners-Lee, “Links and Law” <http://www.w3.org/DesignIssues/LinkLaw> (last accessed 6 September 2002). [31] At <http://www.w3.org> (last accessed 10 September 2002). [32] WC3 Intellectual Rights FAQ, para 4 at <http://www.w3.org/Consortium/Legal/IPR-FAQ-20000620.html#link> (last accessed 10 September 2002). [33] Supra note 13. [34] Supra note 30. [35] See eg Keljob c. Cadremploi (25 May 2001) Cour d’Appel de Paris, 14ème Chambre, Section B, RG No. 2001/03108. Available at <http://www.foruminternet.org/telechargement/documents/ca-par20010525.pdf> (last accessed 20 September 2002). [36] Supra note 30. [37] See eg Havas Numérique et Cadres On Line c. Keljob (26 December 2000) Tribunal de Commerce de Paris, RG No. 2000/099579. Available at <http://www.foruminternet.org/telechargement/documents/tcom-par20001226.pdf> (last accessed 20 September 2002). [38] Supra note 3. [39] See eg NPR <http://www.npr.org/about/termsofuse.html> (last accessed 20 September 2002).