Google Street View - Up Your Street?

Many of my friends and colleagues have been having fun with Google Street View since it went live in the UK last week. My social networking Friends lists are full of people exclaiming "OH MY GOD that's my house!!!" or happily pointing out their car, their garden and in one case, their boyfriend leaning out of the window. Those who live in cities not yet rolled out, bemoan their luck and count how many yards they are from the Googlezone.

Others are not so happy. Privacy International, a well respected privacy watchdog, have already announced their intention to take Google to court on the grounds that they are breaking data protection law, and have made a formal complaint to the Information Commissioner.

Says the Beeb, "Privacy International wants the ICO to look again at how Street View works.

"The ICO has repeatedly made clear that it believes that in Street View the necessary safeguards are in place to protect people's privacy," said Google.

Privacy International (PI) director Simon Davies said his organisation had filed the complaint given the "clear embarrassment and damage" Street View had caused to many Britons."

So is G. Street View ("manic street features" as another BBC piece gleefully calls it) the greatest free of cost and publcly available innovation to hit online mapping ever, or another piece in the jigsaw of ubiquitous commercial and government surveillance in the UK?

Pangloss admits to have been far more excited than worried when she first got the news. Google have invested a pretty large amount of effort into protecting privacy, having learnt from earlier protests and roll outs in the US as well as accepting the reality of ldata protection law in Europe. Faces and number plates have been, with some fairly low margin of error pixelated out. There are indeed errors: we have already had reports of people asking to have maps taken down because they depicted them being sick outside a pub or visiting a well known brothel. But Google have also provided an extremely easy to use take-down request system. Have they done enough?

My esteemed colleague Ian Brown of the OII doesn't think so (and repeated these feelings during a brisk debate last night at a post privacy conference dinner :) Said Ian to the Beeb:

"They [Google] should have thought more carefully about how they designed the service to avoid exactly this sort of thing."

Dr Brown said Google could have taken images twice, on different days, so offending images could have been easily replaced and protected privacy better.

Google says it has gone to great lengths to ensure privacy, suggesting that the service only shows imagery already visible from public thoroughfares."

There are a number of ways to frame this debate. One is the question of opt in to privacy, versus opt out. In the same way that Google Library has tried to push copyright discourse from opt in - consnt by authors to copying of their work - to opt out - asking to be left out of the scheme if not wanting copies to be made (and failed - given the recent settlement?) - it is arguably trying to do the same with privacy here.

If privacy is indeed a fundamental human right, then it can be argued that in principle no one should have to be exposed to even a low risk of an intrusion of privacy by error (let's leave the debate on what that exactly is, plus the debate on how far your privacy stretches in a public place, aside for the minute) and then have to request take down; instead`they should always be asked to give consent a priori. This is probably in gist PI's argument as to why what Google is doing is illegal.

In strict law, Pangloss is not really sure if this is right: the UK DP Act (and the EC DP Directive) do not always demand consent to processing of personal data - there is a well known exception which allows processing to be undertaken without consent if it is in pursuance of a legitimate aim of the data processor (Google) and does not at the same time unreasonably prejudice human rights (DPD, Art 7(f)).

A "few dozen" requests seem to have been made for take down, according to the BBC. If we knew how many views there are on GSV we could work out what percent have been privacy invading.It is probably a very very low percent. But is this the right way to construct the Art 7(f) balance? or should we be looking only at the degree of privacy invasion suffered by each individual data subject concerned - how much they lost - their wife, their job?

We need a real debate here about whether privacy invasion should be regarded as purely an individual issue or a societal problem; similarly whether GSV brings advantages to society as a whole (surely?) and do these outweigh the privacy loss to the few individuals. If GSV sparks this debate it will in itself have been of value.

Ian's compromise solution above - essentially, get it right the first time so as to minimise privacy intrustions requiring post factum take down - is a pragmatic one but does not in essence meet the above theoretical problem. It raises another pragmatic problem too - Google has already spent vast amounts providing a fantastic service for free to the UK public. Yes, they wil gain from ad revenue - but this is still an enormous free gift to the public as a whole. How much more money would it be reasonable to ask them to spend to meet the needs of the very few?

Taking two pictures of every location would presumably have doubled costs. Would fewer cities then be rolled out? Would there be more social and digital exclusion? Will rural areas ever be included in fact? and would someone living next to a person who had had "his" street view pulled out by justifiably irritated at his social exclusion? Should the invaded privacy rights of a few be allowed to stifle technological innovation for everyone? If we consider the P2P debate where the same issue arises - should theeconomic interests of the few in the entertainment industries be allowed to stifle useful innovation for the rest of us? - then generally the informed answer is no. There are many more societal cost/benefit balances to be thought about here.

In the meantime, Pangloss is going to go off to yet another workshop to talk about privacy and trust in next generation networks. Do we indeed trust Google to know where we live and to respect our privacy? Most do but some don't, it appears. Yet Google cannot automate, and thus provide at reasonable cost, the amazing services it delivers for "free" , unless we all agree on this in adavance, or at least are presumed to agree, subject to later opt out. This may be becoming a key problem of the digital era :)