privacy
/ PRY-vuh-see /
Privacy, in the world of AI and data, is the principle that you should have some control over what is known about you and how it is used — that information collected for one purpose shouldn't silently power another, and that being observed shouldn't be the price of participating in modern life. It isn't about having something to hide. It's the same reason you close the bathroom door or seal an envelope: a basic boundary around the self that lets people be free.
Machine learning strains this in a distinctive way. These systems are hungry — they improve with more data — so there is constant pressure to collect everything. And the data behaves slipperily: pieces that seem harmless alone can be combined to re-identify you (your birthday, postal code, and sex together pin down most individuals), and a trained model can memorize and later regurgitate specific examples from its training set, leaking a real person's private text or face. "Anonymized" data is often re-identifiable, which is why true privacy needs more than just deleting names.
Why it matters: privacy is the soil that other freedoms grow in. People behave differently — speak, search, seek help less freely — when they feel watched, a chilling effect that quietly narrows a society even if no one is ever punished. The honest tension is that there is no free lunch: more personalization, better recommendations, and many medical advances genuinely do want your data, so privacy is rarely "all or nothing." It is a set of deliberate choices about what to collect, how long to keep it, who can see it, and what they're allowed to do with it.
Researchers once took a supposedly "anonymous" public dataset of movie ratings and matched it against public film reviews, re-identifying specific subscribers by the unique pattern of films they had rated. No names were in the data — yet individuals were pinned down anyway.
Removing names is not anonymization — unique combinations of ordinary facts can still single you out.
"I have nothing to hide" misframes privacy as personal secrecy. Privacy is structural: it protects journalists' sources, abuse survivors, dissidents, and ordinary people from having data misused later by parties they never agreed to. The harm often shows up not now, but when the data changes hands.