By Atilio Grimani *

First of all, I must clarify that the purpose of this article is essayistic. In other words, its goal is to elicit an opinion from readers, be it in favor or against the tenets that I shall express below**. That said, and if the reader accepts, then there is now a sort of contract governing our relationship while reading this article, which begins below.

Just as the Internet should not be confused with Cyberspace, neither should Google be confused with the Internet (the same applies to any other search engine). This might seem insane to most users. However, the Internet as we know it contains only 4% of existing data. What we see is merely the tip of the information iceberg; that is, information that is indexed by search engines like Google and Yahoo. The rest of the information, the non-indexed 96%, is hidden under the surface. It’s what’s known as the Deep Web. This sub-world, which is unknown to regular users, contains giant squids, white whales and shipwrecks. An impromptu list should include, but not be limited to: databases, movies, music, books, drugs, child pornography and murderers for hire.

This great mass of information is not accessed the way we are used to. To access it, we need special tools, like TOR (The Onion Router) and sites like The Internet Archives or The Silk Road, which in some cases replace the comfort of large Internet search engines with manual searches. Moreover, many of the domains that can be found in the Deep Web do not have the traditional structure with which we are all so familiar; instead they look something like this: (32743grghjbf7438gyfd3g7g378g32.onion).

Thus, it is important to make two clarifications. The first is that even though accessing this data is complex, it is not impossible. The data is still out there, regardless of whether or not it is indexed by search engines. The crumbs that make it to Google and Yahoo are scattered on the bottom of the Deep Web, floating in limbo. The cost of searching for unindexed information is much greater, which may discourage those who wish to obtain that information but are not willing to handle the risks. This notion reveals the tension between the right to access information and, consequently, freedom of expression on the Internet. In the words of Carlos Cortés (here), “This means that, even though personal data may be online, it is not easily accessible. To use an analogy, fishing with a net is one thing, but fishing with a rod is something entirely different: from that point on, whoever wishes to obtain someone’s data will need precise information for doing so. In that same vein –perhaps aware of the tension between the rights involved– the Spanish Agency argued that search engine indexation does not constitute freedom of information. Otherwise, a decision of this sort could involve censorship.” However, these sorts of decisions are hard to translate in our region, because they are not likely to pass the tests of the Inter-American System of Human Rights.

The first conclusion results in a second conclusion with respect to when search engines are ordered not to index certain information (whether in the context of a judicial order on the right to forget or responsibility of third parties for contents regarding celebrities): the measure lacks the efficiency alleged by its defenders. Removing content from search engine index tables does not eliminate its content, it just makes it difficult to find. It condemns information to ostracization, if you will, but this does not prevent someone from finding that information through the above described means. Oblivion is not achieved through these sorts of measures. The Internet does not forget as easily as is thought. In the above cited article, Carlos Cortés also claims that, “This implies that trying to counter [information] or dispose of it (delete or change it) does not only depend on the affected individual or on the third party intermediary involved. Getting the Internet to ‘forget’ such information requires a lot more than pressing a key.”

What I would like to sustain in this article is that measures such as those of the European Court or a measure that could potentially violate freedom of expression standards –such as the decision of the Supreme Court of Argentina in the Rodriguez case– have two underlying problems. The first of these problems is the moral irrelevance of the order to remove data from search engine indexes. This is so because if the goal is forgetting someone’s information –and by forgetting I mean full oblivion of information on a certain topic-, then such orders are absolute failures. A weakened effect, which is seen only in some cases, does not fill the moral debt that the original order attempts to satisfy.

The second problem is that it falls into practical hypocrisy. Users, who might legitimately wish to access such information, are condemned to resort to engines that navigate the darkness of the Internet. Such obstacles violate the right to information of users, as we have seen, but also force them to conduct searches with engines where they could also find various undesired information (from Snuff to hired killers, according to Web myths). In addition, this creates an elitist exclusion of those who wish to have access to information, since only people with additional technical know-how will be able to do so.

Hiding data under the rug clashes with what we know as contextualization. According to Carlos Cortés, who I promise to quote for the last time, “The idea of contextualization (or re-contextualization) can be seen as a critical response to Mayer-Schönberger regarding the issue of partially remembering; however, it can also be seen as an additional tool to expiration dates. Behind this proposal underlies the argument that if an adequate amount of information is available, then users and third parties with access to information will be able to understand and develop their own view on the meaning of that information.” If we toss that information into the Deep Web, the picture will be partial and incomplete.

Developing these points in depth would take more time and space than that available for this article. However, I think this initial assessment of the issue can be useful for sparking the discussion and approaching potential –non hasty– solutions to the problem.

* Worked in CELE and ADC as a researcher and lawyer specializing in digital rights.

** I would like to thank Eduardo Bertoni and Emiliano Villa for their comments and contributions.

*** Translated by Paula Arturo