Your personal data is valuable, which is why people repeatedly try to obtain it illegally and misuse it for their own purposes. We need to be technically prepared for this type of data crime, and we are. We use measures designed to stay one step ahead of digital attackers. For example, through our Cyber Defense Center. We also use methods that de-identify your data so that third parties cannot do anything with it. Read on to learn exactly how we do this.
Sometimes we need to work with your data, for example, to compile statistics or improve services. Before we do, we de-identify your data so that third parties can no longer link it to your identity. We call this de-identification. There are two main methods: anonymization and pseudonymization. It sounds as complicated as it is. So here is the short version:
Once your data has been de-identified, neither third parties nor we ourselves can ever trace it back to your identity. Put another way: After de-identifying your original data, we throw away the key to it. We then work with fully anonymized data sets.
More information about anonymization is available in the video.
We use anonymization when we want to learn about the behavior of large groups of users. For example, to see where people make a lot of calls. This helps us expand our network where it is needed. Another example: Our Park & Joy app helps you find available parking spaces using your smartphone. The service is based in part on anonymous mobility data from the mobile network. This way, you benefit from other users.
Unlike anonymization, this type of data masking can be linked back to your identity. Using a special key. We do not discard this key; instead, we hide it very, very carefully—only we know where. Your identity remains unclear to third parties. But even with the key, we do not learn that you are M. Meier. We only know the pseudonym XX assigned to you. And we know that, for example, XX likes to use app Y.
Find more information about pseudonymization in the following video.
We can only offer certain services when we have more than information about an anonymous group of users and can link specific data to an individual user. For example, this is the only way EntertainTV can make program recommendations based on your individual viewing habits. But no matter how useful a service is, you alone decide whether your data may be used for it.
The de-identification methods described above—anonymization and pseudonymization—have one thing in common: Your data is altered so extensively that third parties can no longer read it. But how exactly is the data made unrecognizable? Primarily with these tools:
Omission:
Important information in your data is removed entirely. For example, your name and street address. This means it can no longer be directly linked to you.
Masking:
In this method, crucial elements of your data are masked with other information, for example, the name Tom Smith is replaced by John Public.
Cryptographic encryption:
Cryptography roughly means secret writing. Today, it uses not only letters or characters, but individual data bits. In short: This process is highly complex, and we use it when large volumes of data are generated at the same time, for example in Big Data applications.
We don't want to leave any backdoors open. That's why, for example, it isn't enough for us to purchase our de-identification software solely from outside vendors – as is standard in the industry. Instead, we developed our own and had it certified by respected data privacy experts. This helps us rule out potential security risks and provide the greatest possible protection even before your data is de-identified.
