Appeal Court Amsterdam: are websites responsible for Google's search results?

The Court of Appeal in Amsterdam has issued a remarkable judgement in the field of internet law this week. The Court confirmed a most controversial ruling by the local district court, saying that administrators of a website can -under certain circumstances be responsible for how their texts appear in search engine results, especially when these results lead to false suggestions and therefore cause damage.

The dispute emerged when Google combined several patches of totally unrelated text on a random website (W) so that the search results on the query “company X + bankrupt” resulted in hits that could be read as suggesting that X was bankrupt.

The website that Google used as a source for mining the search terms did indeed contain the words X and bankruptcy, however, the texts on the site had absolutely nothing to do with a possible bankruptcy of X.

X, a car repair company, suffered  lot (reputational) damage and ordered the website to alter their texts in such a way that the adverse association on Google was no longer made by the Google algorithm. The court ruled in favor of the repair shop.

This ruling is notable as the website in question has displayed no remark on X’ bankruptcy whatsoever and does not contain any illegal or damaging content. The problems have arisen purely from the associations that Google makes. However, Google was not subpoenaed or in any other way summoned to be involved in this case.

Yet, the judge found that W has acted unlawfully by failing to modify their website so that Google no longer makes injurious associations. How did this happen?

The Court chose to make a rather mechanical assessment of the requirements of an unlawful act in Dutch law, instead of acknowledging the broader significance of their analysis for the functioning of the internet.

In short, they dryly observed that: the erroneous association of  information in the rankings of Google were damaging; this can be traced back to the texts of W; W was able to change their texts in such manner that the false suggestion would go away, and that W refused to modify their pages even after repeated requests by X. The court also took into account that W is normally quite keen on their Google appearance, as they actively manage the content of their website to let it score well on Google.

W. fears that websites in The Netherlands will now be overwhelmed by requests of companies to modify their content because of "junk Google results". Especially when this implies that user comments on articles or forums must be altered, this is a threat to the freedom of expression on the internet. I think this fear is justified. Perhaps, in this specific case, it was not that complicated to change a word or two in the text, but it is ultimately not in the public interest (of the Internet user) to demand that sites will actually start doing this. If we do, this will have unprecedented ramifications, as the number of possible keyword combinations in search engines is unimited. How can a website ever manage the appearance of its  content on an infinite amount of search results?

It is nor even the website administrator himself that makes these injurous associations in search engines: Google’s algorhytms do that, in combination with the desire of the internet user. Whoever types 'company X' plus 'unreliable', is creating a negative association himself. Google looks for matching words. Perhaps the combination Google makes is the source of misery, as it brings quotes out of their context. Try it yourself: A little creativity with a search engine quickly delivers a similar case.

My personal sense of justice tells me that the car repair company should have complained to Google, instead of the website. However, the role of Google in this case could not be reviewed extensively by the courts because they were not parties to the dispute. The court proceeded to weigh the interests of the car repair shop against the   interest of freedom of expression of the website owner. The latter lost this balancing act, because W could have mitigated the damage by the simple push of a button.

Unfortunately, the more fundamental interest of a properly functioning internet with freedom of speech of its users, was not weighed in this case. Perhaps this interest is underestimated by the judges or insufficiently highlighted in the case.

At the end of the day, the website was not held responsible for both the distortion of their text by the search engine and the damage of the car repair shop. Also, I’m very glad to read in the judgement that websites are not forced to preventively monitor how their content is processed and presented by Google. However, as a result of this judgement, websites can now be obliged to limit or help mitigate the damage of third parties that is caused by the distorted display of their content by search engines.

It is clear that the development of law on the Internet has still got a long way to go. This story will undoubtedly be continued!

Sources: District Court of Amsterdam, May 13, 2009 (Zwartepoorte vs. Schoonderwoerd); Court of Appeal Amsterdam, July 26, 2011 (LJN: BR3418)