Big Data, Ethics and Privacy

As the world careens towards more collection and analysis of data about – well, almost everything – there is growing evidence that the major issue with big data is that of ethics. This is something that a lot of companies have difficulty grasping and therefore requires action by the people through legislation. It also requires action by the companies and other organizations in the form of paying greater organizational attention to it, perhaps in the form of appointing big data ethicists.

Already as a result of big data, existing privacy laws are badly out of date. That’s because they were written before the advent of big data, and therefore could not have anticipated the issues that are arising.

Companies are using big data to profile their customers, employees and anyone else they encounter. One of the difficulties about profiling is that it is often done with reference to a group rather than to an individual. A good example of profiling is categorization based on race. Thankfully, social legislation has pretty well made such profiling illegal.

However, there are other inappropriate kinds of profiling to watch for. For example, one major credit card company has been reported to have limited credit limits on people who shop at certain stores, on the basis that people who shop at those stores often have bad credit ratings. Clearly, they did some serious digging into the buying habits of certain people and then classified them

Profiling is well established in certain industries, insurance being a prime example. If you are a member of a certain demographic, such as teens, you will pay more for insurance. We’ve come to accept this, but then we know that some teens are very responsible and others not so. Little effort is expended to differentiate between them. We’ve come to accept this state of affairs, despite the inherent unfairness. And that’s the trouble with profiling. The more it happens with different kinds of business, the more unfairness will creep into the system. One of the reasons for this historically has been the lack of data to support differentiation.

And yet, big data offers up the opportunity to differentiate – to do the profiling on a more individual basis. Let’s not let business and governments take the easy route and generate more unfairness. Let’s update the privacy laws to place greater control over the way data is being used, and make sure it is used fairly.

In doing so, we need to remember the basic tenets of privacy policy in the digital age. A key principle is that data about an individual should only be used for a declared purpose accepted by that individual. Big data has moved us way outside of that principle, and we need to address that issue.

There is a disconnect between the ability to achieve individual differentiation in profiling and the right to privacy of individuals in the use of data about them. Therein lies the difficulty. A difficulty that exists, but nevertheless needs to be addressed. That’s a tough one. But nobody ever said the digital era would be easy, did they?