I recently attended a workshop on the legal aspects of Big Data and IoT (Internet of Things) that was organized by Impulse in Brussels.
The presenter, Philippe Laurent, had a very clever and clear way of presenting the problems posed by Big Data and I thought it might be interesting to share his presentation and conclusions with you. In particular he proposed a 4-category framework that is easy to understand and will give you essential guidance in determining what you can and can’t do.
4 types of data and responsibilities
Four levels of responsibility can be distinguished depending on the nature of the data and the ownership :
- Level 0: Big Data used for instance for preventive maintenance. The data collected are only those of objects or animals. There are no personal data concerned. In this case there are no restrictions in terms of privacy. The only problem that may occur is an IP problem if you are handling data produced by others. If you are collecting non-personal data and analyzing them by yourself, then you are safe.
- Level 1: this situation happens when you are re-using your own data. There are no IP problem (the data are yours already) but you are retreating this data for new purposes for which users didn’t necessarily give their consent.
- Level 2: web scrapping falls into this category. You’re collecting non-personal data. However that data was produced by someone else. Keep in mind: it’s not because it’s public that it can be reused.
- Level 3: this is the most complicated situation. It typically happens when you are pumping data from an API and using it for profiling purposes. You are both using data that is the property of third parties (hence subject to the acceptance of a contract) AND handling personal data. Be careful !
This framework will help you understand very easily what is your situation and what are the risks associated to it. I see too many startups (especially in IT) which neglect legal aspects. This is extremely dangerous as it jeopardizes their survival and more generally their business model. By the way, legal aspects can be seen as an inherent part of market research (remember the PESTE(L) analysis)Tags: data mining