Passkeys are everywhere. GitHub uses them. Google pushes them. Your password manager probably nags you about them. They're the future of authentication, right?
Well, yes and no.
While passkeys are excellent for passwordless login, there's a critical limitation developers need to understand: don't use them to encrypt user data.
Here's why:
Passkeys rely on the WebAuthn API's PRF (Pseudo-Random Function) extension to derive encryption keys. Sounds solid, but there's a catch—PRF support is inconsistent across platforms. Your users might authenticate successfully with a passkey but fail to decrypt their data because their device doesn't support PRF.
Imagine building an end-to-end encrypted notes app. A user creates an account on their MacBook (PRF supported), encrypts their notes, then tries to access them from their iPhone (PRF might not work). Boom. Data inaccessible. Your support inbox explodes.
The safer approach: Use passkeys for authentication, but handle encryption separately with traditional key derivation methods. Think password-based encryption (PBKDF2, Argon2) or server-managed keys with proper access controls.
This isn't to say passkeys are bad—they're fantastic for eliminating phishing and credential stuffing. Just don't overload them with responsibilities they can't consistently handle yet.
As the ecosystem matures and PRF support becomes universal, this will change. But today, in 2026, treat passkeys as what they are: a robust authentication mechanism, not a encryption key manager.
Keep your auth modern, but keep your encryption battle-tested.
Also, if you want to check out my blogging project, visit Pluma.ink - a clean platform for documenting ideas.
Top comments (0)