On the lazy reporting of a silly story about increases in data breaches
Over the past couple of days the following have all published stories on the fact that data breaches in the UK have “rocketed” or “spiked” by an “alarming” 1000% over the last five years.
These are mostly well-respected news sources, serving either the tech industries or the public sector. All of them report this story as though the news that self-reporting to the Information Commissioner of serious data breaches is a bad thing. I’ve given the links to the stories not because I want to increase their clicks, but to show the remarkable similarity between them. This is not surprising, as they are all picking up on a press release by Imation (ironically, as a non-hack, I don’t have access to it) which was issued following an FOI request to the Information Commissioner. The response to the request showed that, indeed, in 2007-08 the number of breaches reported to the ICO was 79, and in 2011-12 it was 828. But does that really mean that “Data breaches in the UK have increased tenfold in the past five years” as the BBC put it?
The answer, certainly, is “no”.
The reporting of breaches has increased by that proportion. But that is not particularly surprising. As far as I recall the first guidance issued by the ICO on reporting serious breaches was only issued in July 2010. Before that while there may have been an inferrable assumption that serious breaches should be reported, there was not much in the way of clear direction or expectation until relatively recently. This expectation has become much more explicit since the ICO gained powers to issue civil monetary penalties for serious breaches. Now, all major data controllers know that when there is a serious breach of data security it needs to be reported to the ICO (and for telecoms providers, there is a lawful requirement to do so under the Privacy and Electronic Communications (EC Directive) Regulations 2003).
But is it a bad thing that numbers of reported incidents has increased? Of course not. All breaches of data security are to be regretted, and lessons learnt to avoid they don’t recur. But data controllers need to be encouraged to recognise breaches, and put their hands up when they happen. The ICO even considers self-reporting to be a mitigating factor when assessing what action he should take.
I doubt that many, if any of the people writing for the websites I link to above really think that data security breaches (rather than reports of breaches) have increased 1000% over five years. I’m sure their writers and reporters are very busy, and an eye-catching press release makes for easy copy. But these websites (with the execption of the BBC) are important and specialist sources of information. For them to resort to “churnalism” (a form of journalism in which press release…are used to create articles…without undertaking further research or checking) at the expense of common-sense, especially when it might lead to greater reluctance to self-report, is greatly to be regretted.