When data is anonymized—names and direct identifiers removed—AI can still reconstruct your identity by finding unique combinations of your remaining characteristics. A dataset with age, zip code, and medical condition might be theoretically anonymous, but few people match all three, making you identifiable anyway.
Inference attacks occur when an adversary uses AI or statistical techniques to deduce private information about an individual from datasets that have been stripped of obvious identifying details.
Even supposedly anonymous data released by apps, health services, or research studies can be re-identified using modern AI models, making it critical for individuals to understand that anonymization alone does not guarantee privacy protection.
Peri can explain this concept, give practical examples, help you decide whether it applies to your situation, or recommend a journey if appropriate.
Explore related journeys or tell Peri what you're working through.