Anonymization irreversibly removes any data that could identify a person. Pseudonymization replaces direct identifiers with pseudonyms, but allows the process to be reversed with additional information stored separately — so it is still considered personal data under most regulations.
In GENAI-L (Domain 4, section 4.3) this is an explicit privacy risk in the exam guide when using customer data to personalize or train generative AI solutions.
Anonymization vs. pseudonymization is a concept that can show up on the exam. Check the full guide to see it in context.