Over the past 24 months there has been increasing awareness of data protection and privacy, with everything from the Sony Playstation breach to the recent Edward Snowden NSA revelations bringing the subject into the public eye. New European Union (EU) legislation is proposing fines to organisations of €100 million or 5% of their annual worldwide turnover, whichever is the greater - that's quite an incentive for companies to start getting their act together. As usual, when the brown stuff hits the fan, expect developers to get it in the ear - this article serves as an introduction to developers about "DATA PROTECTION and PRIVACY legislation" and what it means for them.
This article is primarily focused on EU legislation, although having said that, it should be useful in general no matter the location. In the US for example, many organisations have signed up to "Safe Harbour" - this means that they must implement the same protections as their EU counterparts in relation to data, if they deal with the data of EU citizens. An increasing amount of countries are implementing new privacy legislation, and the general principles in this article would apply to most of these.
My own development background has focused the past 10 years in the Healthcare sector - as you can appreciate, this area is more concerned than most with protection of data and patient/consumer privacy. My interest in data security and need to know more led me to eventually getting certified in this area and I am always amazed at the lack of knowledge of the area - hopefully this will go a small way towards filling the gap at the frontline of data gathering!
I am not going to go into great detail on specifics, if you need this, there are many great resources waiting to be Googl'ed (or Bing'd!). At the end of this article, I give links to the website of all of the EU data protection offices for further detail. The main focus of this article is to discuss each area of the over-arching legislation in relation to how it might affect developers and operations.
Data protection in EU countries is based on what's known as a "directive" - in this case, its 95/46/ec. The directive is prescriptive in some areas, and in others leaves things up to the interpretation of the national government. This means that in general, data protection and privacy legislation across the different countries of the EU is more or less the same, with some local differences. It's difficult to get things perfectly right in all jurisdictions, but if you start with the core rules, you are most of the way there.
#1 - Obtain and process information fairly
#2 - Keep data only for specified, explicit and lawful purposes
This rule says that you can only keep the data you collect, if you use it for the reasons you told and agreed with the user (competition entry versus marketing spam), you inform the user clearly about the information you are gathering, and critically, you need to know exactly what kind of data you are gathering, and where it is stored. For developers, this opens up questions such as defining the type of data gathered, and keeping careful track of where it is stored, who has access, etc.
#3 - Use and disclose only in ways compatible with the purpose
What this means is that how you use the data gathered, must be strictly in accordance with what you told the user it would be used for. This rule reiterates the example of "prize draw versus marketing spam". Disclosure is a big part of this rule. It states that you may not disclose the information received, to another party / organisation, unless you have the *express permission* of the user. There are clever ways marketing types try to get around this, but really, it's best to do the right thing - users are getting quite tired of privacy abuse these days and it does more damage than its worth to share data in an unauthorised manner. For devs, the thing to watch for here is that you lock down data, and only allow access through for example APIs that access control has been implemented.
#4 - Keep data safe and secure
Oh dear - this is a can of worms! ... this moves beyond the realm of the developer and now involves your IT infrastructure experts. This is a specialised area and you should take advice where necessary. From a developers point of view, you need to ensure data is stored encrypted both at rest (in files, in database, etc.), and in transit (https everywhere!). Depending on the sensitivity of the data, you also need to ensure strong access control is in place, with PKI and two factor authentication where necessary. Careful records should be kept, and regular audits of security need to be carried out, and a 'designated person" needs to be assigned to be responsible for data/information security.
#5 - Keep data accurate, complete and up to date
Mmm, up to now, this data protection stuff seemed reasonable to keep on top of - so here is where we start to go down the rabbit hole.... from a development point of view, how can you help? .... the easiest way is to ensure that all data captured is date-stamped, so that you can set up time based audits of information, that checks when was the last time (a) the data was confirmed with the user for accuracy (a refresh of data every 12 months is generally expected. eg: has the user changed address in the past 12 months? has their marital status changed?) (b) if the data has not been updated that is flagged as a problem and action taken.
#6 - Data held must be adequate, relevant and not excessive
More of it .... this is a very fine balance between what the organisation wants to know about a user, and what it absolutely needs to know *at a bare minimum* about a user. This rule must be taken in context with #1 ad #2, that is, the data held must not only be the bare minimum you need to know about the user, but the minimum you need to know in the context of the reason you took the data in the first place. This means for example that if we hold a competition, and gather user data in relation to that, unless there is a reason for asking and storing the users age (for example if the price required adult status), then you may not ask the question or store the data. Of course, again, the marketing folk can put in terms and conditions that require the information, and the user can of course choose not to enter, but they are walking on very thin ice. From a development point of view, when building solutions that gather and store data, you should ask what the reason is for the data being gathered and what justification there is for gathering it and using it.
#7 - Only retain data for as long as is absolutely necessary
This one is straightforward, but can be a bit tricky as other rules can be contradictory. This is where you need to take local legal advice if unsure. Let's look at an example - if you apply for a job, and you don't get offered the position - should the potential employer be allowed to keep a copy of your resume? ... you may say now, however, they may argue that you may come back and sue them for some kind of discrimination for not being offered the position so they need to keep it ... fine, but for how long? Under employment and contract legislation, it may be required that all application documentation, plus any notes and supporting decision making documentation and communication be kept for a prescribed period of time. In this case, the employment or contract legislation may trump the data protection and privacy legislation. For a developer, the best thing to do is to keep track of what data you have, where it came from, and link it back to how long it should be kept - this can be facilitated by timestamping and tracking data in and out.
#8 - Give a person a copy of their data on request
So, last but not least, data access requests. This rule simply states that you must give a person a copy of any and all data you hold on them, subject to certain rules and exceptions. Rules/exceptions generally relate to medical and legally sensitive information. This rule may seem simple, but its implications mean that someone (that's you developer person!), must keep track of what data is in the system, who it relates to, where it came from, what it is being used for, etc. etc. so that, when a person/user asks for a copy, you can readily access it and hand it over. Data access relates not only to record based data but also to any video/audio, etc.
The data legislation (very) briefly outlined above is about to change. A process of consultation has been going on for the past while and new updated laws are due to be in place in the next 12/24 months. These rules bring the legislation up to date and into line with modern data handling, and encompassing things like social networks, cloud storage etc. While up to now actions taken by data protection and privacy controllers have been minimal, they are ramping up to take on the new rules and there is a lot of talk of heavy fines coming down the line. It is important that developers are aware of the rules, and take action when designing and building out systems to ensure that the organisation they work for have technologies and systems in place to enable them to adhere to the legislation. Hopefully this article gives you some tips and pointers on how to get started. I also attach a PDF from the office of the privacy commissioner in Hong Kong that has some useful pointers especially for mobile app developers.
Ref: Links to national organisations responsible for data protection
- 21/12/2013 - Version one published