The short answer
Health and genomic data are highly sensitive under the GDPR. Processing typically needs a specific legal ground, such as explicit consent or a defined research exception.
General orientation, not legal advice.
Why this matters now
Health data sits under stricter requirements. In Germany, further rules in the BDSG can apply, for example for research and archiving in the public interest. For a technical team, the legal ground, the operating model and the safeguards have to fit together.
What to do
- Name the legal ground, such as explicit consent or a research exception.
- Settle an external provider as a processor under Article 28: contract, instructions, technical measures, evidence.
- Review a transfer to a third country during architecture, with a risk analysis and guarantees.
- Where the risk is high, check whether a data protection impact assessment under Article 35 is required.
What this means technically
Pseudonymisation lowers the risk. It does not remove the GDPR duties. Full anonymisation of genomic data is often hard to sustain, because re-identification still has to be assessed. Lowering the risk and the duties falling away are two different questions.
Common misunderstandings
-
Pseudonymisation cancels the GDPR duties.
It lowers the risk. The duties remain.
-
Genomic data can simply be fully anonymised.
Full anonymisation of genomic data is often hard to sustain. Re-identification risks still have to be assessed.
Where to start
With the question of which legal ground, which operating model and which safeguards fit together. If you want support with that, you can ask for an assessment. It is not legal advice.
As of 15 November 2026
Not legal advice. As of 15 November 2026. A binding reading needs specialised data-protection and legal advice.