Copy Control by Eric Goldman and Shawn Molodow

Losing Control of Information on the Internet

by Shawn N. Molodow and Eric Schlachter


The judge orders your client to remove offending materials that they have published on the Internet.  Your client explains that they have already removed the information from their website.  Hasn’t your client done all that they can?  No, the complainant responds — even though your client has removed the offending material from their website, the offending material is “all over the Internet.”  A search with your favorite Internet search engine confirms this; there are more citations and copies of your client’s materials scattered across the Internet than you have the time or patience to read.

If your client has removed the offending material from their website, how and why is the information scattered across the Internet?

The answer lies in how information is organized on the Internet to reduce congestion and speed-up retrieval of information.  The Internet connects computers from all over the world, and under optimal conditions, such connections may be at near the speed of light.  However, operation of the Internet is rarely optimal.  Not unlike automotive freeways, the Internet gets congested when there are too many users or choke points arise or major connections go down or a variety of other problems arise.

In order to reduce congestion, copies of information made available from one server are stored in other locations.  For example, to avoid trans-Atlantic bandwidth constraints, an exact duplicate of a California website may be also made available on a Europe-based server in a process called “mirroring.”  Mirroring allows information to be stored geographically closer to the user, which reduces the time, distance and congestion caused by retrieval and delivery.  Another example is proxy caching, which involves storing frequently-requested web pages on a local server so that when a user requests a web page, the locally stored copy is delivered rather than retrieving the information fresh from the Internet.  A third example is search engine directories, which often copy information from websites and store the information to facilitate faster and more efficient searches later.  In the case of full-text indexing, an entire copy of a document is stored in a full text or nearly full text database which is searchable for key words or concepts.  A fourth example is archiving, which refers to a database of historical information, such as in the case of newsgroups and mail lists where the information would otherwise be evanescent.

Because so many copies can be made for different purposes, by making information available on the Internet, the original publisher of the information could lose control over its information.  This loss of control can raise some difficult legal issues.  For example, the process of copying raises tricky issues under copyright law, which has triggered a debate currently raging nationally and internationally.  Also, advertising-driven websites are frustrated to learn that proxy servers collect the “hits” and other user data that are prized by advertisers.  Finally, harm could be created when the remote copies persist despite changes to the underlying pages.

A recent case, Toys ‘R’ Us v. Akkaoui, 1996 U.S. Dist. LEXIS 17090 (N.D. Cal. October 29, 1996), illustrates how the latter point — the loss of control over harmful persisting copies — can create difficult situations.  In August of 1996, Toys ‘R’ Us learned that a shopping service featuring a variety of sexual devices and clothing was operating at the domain name    Toys ‘R’ Us sent a letter demanding that Adults ‘R’ Us immediately shut down their Internet site.  Adults ‘R’ Us declined.  Toys ‘R’ Us then sued for trademark dilution and infringement, false designation of origin and unfair competition.  Toys ‘R’ Us also sought a preliminary injunction.

To determine whether or not to grant a preliminary injunction, the Court focused on the likelihood of success on the merits and whether Toys ‘R’ Us would suffer irreparable harm.  The Court’s analysis of the balance of hardships on the parties was brief, probably because Adults ‘R’ Us did not inform the Court that a preliminary injunction would impose an undue burden.  As to the merits, the Court found that the Toys ‘R’ Us family of marks were famous and distinctive and that “Adults R Us” tarnishes the “R Us” family of marks.  Accordingly, the court concluded that Toys ‘R’ Us was likely to prevail on the trademark dilution cause of action.  As to irreparable harm, the court held that continued references to Adults ‘R’ Us would likely lead to irreparable injury to the “R Us” family of marks.

Adults ‘R’ Us argued that there was no need for an injunction because they had removed the allegedly offending material from their website.  Toys ‘R’ Us countered that sites identified as Adults ‘R’ Us could still be found on the Internet.  Most of these sites were no longer hypertext-linked to the Adults ‘R’ Us catalogue of sexual products, but some still were.  Understandably, Toys ‘R’ Us wanted the name Adults ‘R’ Us more completely removed from the Internet.

The Court’s injunctive order included requiring Adults ‘R’ Us to immediately discontinue the use of domain name “;” surrender all advertising, products and related materials that bear the name or mark Adults ‘R’ Us or a similar mark, and cancel all orders for such items.  The order continued that Adults ‘R’ Us was directed to “notify in writing and direct all publishers of directories or lists, including Internet search engines, in which Defendants’ “Adults R Us” name appears, to delete all references to this name from their public databases, search directories, directory assistance . . . and to delete all forwarding or “cache memory” or storage mechanisms referencing such name . . . .” (emphasis added).

The requirement that Adults “R’ Us notify all publishers of directories and lists on the Internet creates an impossible situation.  As already described, there were many different places where Adults ‘R’ Us information had propagated throughout the Internet.  There are millions of servers connected to the Internet; many of these servers could not possibly be found by Adults ‘R’ Us, regardless of the diligence of their search.  Furthermore, even when Adults “R’ Us has complied with the injunction and notified a directory or list, it may take a long period of time before the reference is actually removed–for example, in the case of a full text index database such as Alta Vista, it could take months for a request to remove information to be honored.

The expense and effectiveness of Adults ‘R’ Us actions to comply with the Court’s order are unknown, but the lesson is clear: archiving, caching, mirror sites and search engine indexing create potentially problematic scenarios for Internet publishers.  These storage mechanisms may have already created legal “time-bombs” lurking on the Internet that will be felt for many years after an Internet publisher has shut down or changed its site.

The Adults ‘R’ Us case is a vivid reminder that the technology of the Internet is different from historical antecedents.  Because a company faces loss of control over its information, there are new dilemmas that require careful planning and foresight.  Perhaps the best lesson from the Adults ‘R’ Us case is that, before becoming an Internet publisher, a company must have an exit strategy well in hand.


Shawn Molodow is an attorney practicing in the Information Technology Group of Cooley Godward LLP.  His practice is split between advising clients on licensing related issues and general corporate matters.  He can be reached at

Eric Schlachter is an attorney practicing cyberspace law in the Information Technology Group of Cooley Godward LLP.  He is also an adjunct professor of cyberspace law at Santa Clara University School of Law.  He can be reached at