Michel pointed me towards an interesting article on Google and it’s searches:

Google started to filter search suggestions that include terms associated with copyright infringement like “torrent”, “bittorrent”, “rapidshare”, “megaupload”. It’s a slippery slope and Google’s suggestions will be less useful since they’ll no longer include many popular searches.

Last month, Google explained that this is one of the changes intended to address copyright infringement. “We will prevent terms that are closely associated with piracy from appearing in Autocomplete. While it’s hard to know for sure when search terms are being used to find infringing content, we’ll do our best to prevent Autocomplete from displaying the terms most frequently used for that purpose.”

Blacklisting keywords like “torrent” is a terrible way to prevent copyright infringement since users can always type queries without Google’s help. The main consequence is that Google will appear to be broken and users will no longer trust the suggestions because they’re censored. … As Mashable says, “this is a subtle form of censorship, and at first glance it seems trivial. However, even though the censorship is slight, it still indicates Google’s willingness to change its search protocols to satisfy the needs of a certain business group, in this case members of the entertainment industry.”

It is key to understand that Google is not filtering searches per-se, but filtering the auto-complete suggestions that it automatically generates. So as you type in a word, Google suggests possible search avenues for you by guessing the completed word/s. So how does this look in practice? Supposed I typed in ‘batm‘, Google looks at the characters so far and guesses that I’m probably going to add ‘an‘ to this so suggests searches. This is what it looks like:

.

Now I try ‘torren‘ not yet adding the last ‘t‘ to make ‘torrent‘:

.
This time the partially completed word has no suggestions at all. However as soon as I add the last ‘t’ to the word:

.
You can see that the suggested searches kick in again. The top suggested link is a guide to Torrent sites, the second is to the Pirate Bay. Which brings me to what I think is driving this. During the recent Pirate Bay trial, the point was made by the defence that what the Pirate Bay did was akin to what Google does – just a search facility. Remember torrent files are neither illegal in themselves nor contain any copyrighted material – they are just simple text files. I think Google is trying to balance the difficult tasks of ensuring that we believe it to be the most authoritative search engine on the net with (legal?) pressure to censor it’s output. If we, as users, no longer trust Google to return accurate results for a search – then it’s main reason for being is doomed.  (Interestingly Stack Overload’s Joel Spolsky made this point on the Guardian Tech Podcast, that Google has to return accurate searches above even it’s own commercial interest…) But if it does nothing in the face of legal actions, then it could be forced to comply with a ruling that achieves the same result. This I think is their balancing act response to this situation.

It is a slippery slope, but one that slides in many directions…

(Also posted on my blog)