The Processing of personal data is caught between technological progress and strict data protection requirements. The General Data Protection Regulation (GDPR) requires the protection of individual rights, but also allows innovative approaches through Anonymization. The foundation Data protection has published a practical guide to Anonymization published, which provides a sound basis to personal data securely and legally compliant.
Definition and meaning of anonymization
Anonymization refers to the conversion of personal data into a form that no longer allows conclusions to be drawn about an identifiable person. The decisive factor is that the data after Anonymization no longer fall within the scope of the GDPR are subject to.
The Anonymization is often used to be able to use data for research purposes, market analyses or software tests without violating data protection rights.
The demarcation to the Pseudonymization is essential, as pseudonymized data is still considered personal data and is subject to the GDPR fall.
Requirements for effective anonymization
The GDPR does not require a specific method for Anonymizationbut defines requirements indirectly. It is essential that the data is not collected by the controller or by Third can be assigned to a person with reasonable effort. Factors such as costs, technological possibilities and the probability of re-identification play a central role here.
Inspection obligations:
- A responsible party must be able to prove that the Anonymization is practically irreversible.
- Indirect identification features such as gender, date of birth or zip code must be carefully checked to rule out any conclusions.
Anonymization methods
The guide describes several methods that can be adapted to specific requirements:
- Randomization: Data is alienated by random changes to the values in order to prevent conclusions being drawn.
- Generalization: Values are transferred into larger categories, e.g. the aggregation of age data into age groups.
- Differential Privacy: Usage-dependent biases protect individual data points in aggregated data sets.
- Synthetic data: Artificially generated data that is statistically similar to real data but has no personal reference.
Reading tip: Cookie consent management - secure consent for companies
Legal challenges of anonymization
Processing procedure as part of the GDPR
The Anonymization itself is regarded as Processing personal data and is subject to the requirements of the GDPRincluding a legal basis. Only after completion of the Anonymization the applicability of the data protection regulations to the anonymized data does not apply.
Re-identification risks
Particular attention must be paid to the possibility of anonymized data becoming personally identifiable again through external information. The use of an "attacker model" for risk analysis is essential here.
Integration of third parties
ThirdThe companies and processors that carry out anonymization must comply with strict contractual regulations in order to guarantee the security of the data and the independence of the anonymization process.
Possible applications and examples
The guideline describes four central application classes:
- Anonymization as Deletion: Replacement of the Deletion of personal data by Anonymizationfor example, in the case of application data after the conclusion of a selection procedure.
- Disclosure of anonymized data: Salary benchmarks or sales data are described by way of example, which are calculated according to Anonymization can be passed on in a legally secure manner.
- Anonymization in the training of algorithms: Techniques such as federated learning make it possible to use data without it being centralized or depersonalized.
- Anonymization for software tests: Synthetic data ensure that tests can be performed without access to real personal data can be carried out.
Anonymization offers companies and institutions the opportunity to use data efficiently and in a legally compliant manner. The data protection foundation's practical guide shows that this is not only a technical challenge, but also a legal and organizational one. Responsible persons must not only carefully plan and implement anonymization processes, but must also be able to demonstrate that the measures meet the high requirements of the GDPR are sufficient.
The clear separation between Anonymization and Pseudonymization is crucial in order to minimize legal risks and at the same time exploit the full potential of the data.
German 
English 
Italian 
French