Regulation & safety4 min read

Safety and data privacy

Patient data is the fuel of AI — and protecting it is a legal and ethical requirement woven into every step.

Medical images are among the most sensitive personal data that exist. Their use in AI is governed by strict legal frameworks — in Europe the GDPR, in the United States HIPAA and state laws — and by professional ethics.

Protection is both technical and organisational: encryption in transit and at rest, strict access controls, de-identification before research use and audit logs of every access. Privacy is not a feature bolted on afterwards; it is designed into the system from the start.

How patient data is protected

  • Encryption and strict access controls
  • De-identification before research use
  • Informed, revocable consent
  • Audit logs for every data access

In brief

GDPR and HIPAA govern patient data in AI.
De-identification protects identity in research.
Privacy is designed in from the start, not added after.

Keep learning

Data quality and ethicsAI HubBias and fairness in AIAI HubClinical validation and evidenceAI Hub