Why local encryption changes everything

Where encryption happens decides who can read If encryption happens on the server, the server has access to plaintext at some point. If encryption happens on the client, plaintext never leaves the client.

What this implies for trust With local encryption, you do not need to trust the provider's *behaviour* with your plaintext. You only need to trust the cryptography — which is public, widely reviewed, and does not change behind closed doors.

What this implies for compliance Local encryption simplifies a lot of compliance discussions. A provider that cannot read your data cannot leak the *contents* of your data in the first place.

The single trade-off Account recovery is harder. If you lose your password and your recovery key, encrypted data cannot be recovered. That is the price of architectural confidentiality — and it is the right trade-off for sensitive files.

The trust model is inverted With server-side encryption, you trust the provider, their staff, their KMS, their incident response, their legal team, and every government able to compel them. With local encryption, you trust the math and your own password. The provider becomes a dumb storage layer — exactly what cloud should have always been for sensitive data.

What stops working — and why that is good You lose server-side search, full-text indexing of your documents by the provider, and AI-generated previews. In return, you gain real confidentiality. The trade-off is intentional, and it is the same one every serious zero-knowledge product has made for two decades.

A change of category, not a feature Local encryption does not make a traditional cloud "more secure". It makes a different category of product. DRIVUNO is built in that category from day one — there is no legacy path back to plaintext, no admin override, no escrow.