The best way for me to answer is just to step back for a moment and maybe set the context that Google is a company that is driven by innovation; we're constantly innovating and constantly creating new product. We think that has great economic benefit and great benefit to consumers, and we think this has been proven on a number of occasions.
That being said, the reason that Google retains non-blurred images at all is for product enhancement. So in order to make the product better, Google software engineers keep the photos around for a period of time during which they do improvements on the product and for a number of other reasons. I can describe one product improvement or product enhancement. The product operates at immense scale. So Google is taking, as you can imagine, millions of photos, and there are people who are inadvertently and incidentally captured. Google blurs these images at an incredible scale, and although our software is very good, it is not always perfect, and it occasionally generates a false positive. So occasionally the software will mistake a round sign for somebody's face. For the purposes of trying to improve the software, we keep the images around.
However, Google recently has made a revision to its data retention policy and has decided that it needs to keep the images around for an adequate but non-excessive period of time, after which those non-blurred images will be permanently blurred and therefore rendered anonymous.