Facts 1 of 10: Pseudonymization is not the same as anonymization.
Within the AVG, pseudonymization is defined as “the processing of personal data in such a way that personal data can no longer be related to an individual without the use of additional information.” “This is under the condition that this additional information is stored separately and is furthermore subject to technical and organizational measures that ensure that personal data are no longer related to an identified or identifiable natural person.” This means that the use of additional information can lead to the identification of individuals, which is why pseudonymized personal data still falls under personal data.
At the other end of the spectrum are anonymous data that is not traceable to specific individuals. When data is truly anonymous and individuals are no longer identifiable, this data does not fall within the scope of the AVG.
Is anonymized data always anonymous data?
To do this, we first establish the origin of the data, in addition, we assess a number of phases within the information life cycle: Collect, Use, Provide and Store. For each phase, a number of weighting factors lie around simple questions such as: Has consent been sought and granted? Down to the final phase of the Information Life Cycle: delete the data.
Irregularities in these parts are common, mostly unconscious. For example, permission is not always sought or granted. Usually this is done on the basis of generic announcements on a website. This does not disclose the intended research objectives. Data are kept in 4 to 10 different places. This is not always known, nor is it always clear where the data is stored and whether it is secure. Thus, “deleting” data can lead to latent backups.
How do I determine if data is covered by the AVG and may be used?
A better question is: What are the risks and how do we set up a careful process from start to finish? There are methods and techniques for that. Generally, for anonymization, there are two options: Synthetic data generation and Data masking. The latter uses functions such as Shuffle, Scramble, Blank etc. Masking is all about reusing data in databases. Data generation creates new data that did not already exist. A third method is a hybrid approach: Generate synthetic data as a masking technique to replace existing data.
Anonymization is difficult and is rapidly becoming more difficult due to technological advances. There are good examples of processing anonymous data. A fairly unique successful example is the Central Bureau of Statistics (CBS). By law, CBS is designated as an organization to produce statistics for policy and research. CBS is normative when it comes to applying techniques for anonymizing data.
Discuss your situation and learn more? Make an appointment!