“I have nothing to hide” — the wrong lens for AI privacy

“I have nothing to hide” is the wrong lens for AI privacy, because it confuses privacy with secrecy. Privacy isn't the art of concealing wrongdoing; it's control over information that belongs to you — and, most often when you use AI at work, to others: your clients, your patients, your colleagues. That control isn't justified by guilt; it's a default stance.

Privacy isn't the same as secrecy

Legal scholar Daniel Solove dismantled the “nothing to hide” argument: reducing privacy to concealment means accepting that only wrongdoing justifies protecting yourself. Yet you close the bathroom door with nothing to hide; you seal an envelope without a plot. Privacy is the right to a closed door — not proof that you've done nothing wrong.

It's almost never only your data

At work, what you paste into AI is rarely your secret: it's a client's name, a patient's file, a colleague's salary, a partner's contract. “I have nothing to hide” doesn't hold — because it's not yours to decide to hand over someone else's data. You're its custodian, not its owner.

• A client's name isn't your information to give away.
• Health data concerns the person involved, not you alone.
• A trade secret belongs to the company that entrusted it to you.
• Even “harmless” data, cross-referenced with other bits, can re-identify someone.

Reclaim the default reflex

We used to have a simple, unspoken reflex: don't hand what's private to just anyone. AI broke it, because it's useful and pasting is fast. Reclaiming it doesn't mean giving up AI — it means deciding, by default, that sensitive data doesn't leave in the clear.

That's what ONYRI Sanitize does: it detects sensitive data and replaces it with tokens before sending, then restores the answer in your browser. You keep AI's power and control over what was entrusted to you — the goal, really, is to rediscover something essential we'd quietly forgotten: our privacy.