Why deep linking can be dangerous

Deep linking has become the latest hot topic in e-commerce. In the not too distant past Web site owners would have been...

Deep linking has become the latest hot topic in e-commerce. In the not too distant past Web site owners would have been delighted with any free publicity that links to their sites might have created. Now, they are quick to threaten injunctions in order to prevent their content being accessed by an unauthorised link. This reflects...

the huge amount of time and money businesses now invest in their Web sites. A link to a business-critical deep page of a Web site can be a very serious commercial matter. The link may, for example, bypass revenue-gathering pages or make use of a page from another site which has particular value, or which was difficult or costly to produce. The creation of a deep link, as such, is not the problem. A link does not copy a page from another site. It merely redirects the user to that page. The browser, or user, will make a temporary copy of the page in the process - but that is another matter. One of the earliest cases on deep linking was that of Shetland Times v Wills. That case is sometimes cited as legal authority for the suggestion that deep linking will always be illegal. But that conclusion is far from accurate. The difficulties arise, in a legal sense, when a deep link or frame somehow creates a commercially sensitive intrusion, for example, by way of infringement of copyright, of database rights, trade mark infringement or what may be interpreted as "passing off". Some of the best examples of deep linking cases come from Europe. In the case of Stepstone versus OFIR, the German courts granted an injunction to prohibit a link to the claimant's Web site, where a rival recruitment agency had created the link. That link enabled the rival to display, on its own site, specific job vacancies from the other site. The link was obviously unauthorised. The court granted the injunction - but not just because the link was a deep link. The basis of the decision was that, in the process of creating the link there had been an infringement of database rights which the agency was entitled to protect. Database rights are recognised in England as well. The recent case of William Hill v British Horse Racing Board related to whether William Hill had taken data derived from the board's database to use on its Web site. In finding in favour of the board, the court commented that database rights would protect virtually all collections of data in searchable form. That decision is of major importance to the concept of deep linking, as in reality, most links tend to be to pages containing useful data. The courts will, however, protect these valuable assets. There are a number of misconceptions about linking. For example, it is commonly thought that a link to a home page will always be permissible. Generally, a deep link is liable to be more problematic than an ordinary link to a home page - but this fact alone will not be determinative. In direct contrast, it is often also assumed that, by publishing any page on the Web there is an implied consent given to create a link to that page - whether or not it is the home page. While there seems to be acceptance that there is an implied consent to link to a home page, the same cannot automatically be said for deep pages. The situation is not helped by the existence of contradictory case law in different countries. Although some German case law supports the principle of a general implied consent, the better view is that deep linking will not be permissible if it infringes some other right, for example copyright. Preferably the link itself should make it clear that the page to which the link is made is, in fact, from a completely unrelated site. A Danish case gives a good example. There, music copyright owners have obtained compensation from individuals who had created links from their own home pages to unlawful music files published on unrelated sites on the Internet. The link in that case constituted a copyright infringement tantamount to an unlicensed public performance. The US courts have considered the use of spiders to retrieve information which was used to compile lists of deep links to Web pages in the context of trespass. The claim was successful because the search system was found to have the potential to create an unacceptable interference with the claimant's site, which amounted to a trespass. The inclusion of a linking policy or disclaimer will generally not be enough to resolve all disputes - disclaimers do not provide wholesale immunity. So although the Internet does exist to provide a free exchange of information and ideas, that does not mean that Web sites can be plundered for content. Michael Clinch is head of the litigation department at law firm Picton Howell

Read more on IT risk management