On 13 May 2014, the Court of Justice of the European Union recognised the 'Right to be forgotten' ruling against Google and all other search engines. You probably don't need to be reminded, but it means that if inaccurate or out of date material (or even just something that someone doesn't like that is said about themselves) is found in Google, that person can apply to the local variant of the engine to have it removed - or rather 'delisted', which is to say that the material itself on the website would still be there, just that Google wouldn't point to it with a direct search.
Google made it fairly easy to overcome this however by simply redirecting the browser to the .com version of the search engine, where the offending article could still be easily found. More recently they have made it slightly more difficult to get to .com from the local version by reducing the number of links on a local Google home page. However that's not enough for CNIL, which is the French data protection authority. CNIL has requested Google to carry out the delisting of several results. It was expressly requested that the delisting should be effective on whole search engine, irrespective of the extension used (.fr; .uk; .com …).
This is opening up a real can of worms, because it's really implying that any national court has jurisdiction over the entire world. If one court finds that something is inappropriate, then Google's work around of removing the material in that country will no longer work. Now, is this censorship? Depends on your viewpoint I think. It's not censorship in that the material is still available, just much harder to find - we're back to where we were in the days of print only; you have to wade through a lot of data to get what you need. On the other hand, given the amount of data that's now available, if material has been 'delisted' (which is a far better term that the clumsy 'right to be forgotten') and that makes it virtually impossible to find - is that a de facto form of censorship anyway? Furthermore, if it's possible to find material that says that 'John Smith is a paedophile' then to all intents and purposes - he is! You had to dig deep into local newspaper archives to find it, and you have that information. However, if it transpired that 5 years later it turns out that he wasn't a paedophile and was unjustly imprisoned, it's just as difficult to find that; probably more so since is a searcher going to spend hours further trying to hunt down information that may not actually exist? If the details of the case AND the later overturning of it are both freely available isn't that more helpful?
We also have the problem of competing claims. If country A demands the global delisting of an article, and country B says that the material should remain available, what is Google to do at that point? (I'm not aware that this has happened yet, but it's going to, I'm sure.) What these national courts are expecting the search engine to do is to act as a global servant of that court, when in actual fact it has no jurisdiction to do so, and that's not its role. This is another problem that I have with the whole RTBF or delisting issue - Google is expected to look at individual complaints and make up its own mind on the validity of each case. Is that really what we want from a multinational? To act as a judge and jury for us?
What then happens when the boot is on the other foot, and something happens that the Americans don't like. Say for example that I've got material on a server that conflicts with the American Patriot Act and I have an office in New York; I'm pretty sure that the Americans would want access to it. They may want to data mine the information or delete it entirely. Whistleblowers are either to be protected the world over, or they're not. However, there's a danger that I'll go off down a side road at this point, so I'll just park this point here and move on.
What can search engines do in this situation? Well, they can decide to complete ignore the ruling and take whatever slap on the wrist they get given. If they really don't like it they could always pull out of a particular country. Google could look at the IP address of an individual and if they are based in France could simply remove the data from the search results; that's relatively easy to do. However, it's also very easy to get around it with a VPN that tells the engine that they're actually not in France, they're in Ohio instead. Alternatively they can just pull all of the data when requested. As mentioned previously, this is going to be a difficult call. Russia is implementing their own RTBF law only it goes much further than the European version, since they'd just have to say what material they wanted removed AND it includes reference to public figures. Does this mean that a search engine is going to have to remove something on Putin that the Russians don't like? I really can't see that happening to be honest.
Now, if that wasn't difficult enough, CNIL actually wants a different solution. They want material removed for ALL French citizens from where ever in the world they are. Quite how Google is supposed to do this isn't quite clear. There are options I suppose - by placing cookies onto a machine which can't be removed by the owner (hey, I didn't say that it would be a sensible option!), by limiting by language (ditto), or limit by registered user (in which case don't search while being logged into your account). None of these options work at all well. It doesn't appear that CNIL has really thought this through in any detail at all, or indeed has any real understanding for how the internet and search engines work - no change there then.
How does that affect information professionals? It's something that we are going to have to keep a really close eye on. We need to continually remember that results from one engine are not the same as you'll get with another, and even from the same engine if you're using different country versions, but that's nothing new. If it turns out that information on John Smith the paedophile is blocked in the UK, and you either know this or suspect it, to what extent should you inform your client of work arounds? If and when the Russian RTBF law comes into being, or indeed that of any other country, do we need to keep lists of which country says what? How much do we cross check from one source to another? Should we start to actively seek out search engines to use that don't have the same high profile as the big ones? Or just point someone in the direction of a computer and let them get on with it themselves?
There are no easy answers, and it's just going to get more difficult. CNIL gave Google 15 days to work something out and that window is coming to an end. It's going to be interesting to see what happens next!