FIDO Alliance publishes new spec to let users move passkeys across providers

FIDO Alliance’s new specifications for moving passkeys and other credentials between providers are prompting both optimism and skepticism. Supporters see standardized export/import as a major usability and security win over passwords, especially for non-technical users vulnerable to phishing and weak-password reuse. Critics counter that current passkey ecosystems are still provider-centric, hard to back up or self-host, risk vendor lock‑in and attestation-based gatekeeping, and remain confusing or fragile for edge cases like multi-device power users or disaster recovery.

Lock‑in, control & exportability

  • Many see the new FIDO credential exchange spec as overdue: export/import should have been simple (like CSV for passwords) and available from day one.
  • Strong concern that big platforms (Apple, Google, etc.) intentionally made passkeys non‑exportable to create ecosystem lock‑in; users often cannot move keys out of iCloud/Google.
  • Others argue the new spec is a net positive: it standardizes what were previously proprietary, opaque export/import mechanisms and should reduce lock‑in.
  • Power users want SSH‑style control: raw key files they can copy, back up to paper, or self‑host (KeePassXC, Bitwarden, Vaultwarden). The inability to access private keys directly is viewed as paternalistic.

Security properties & threat models

  • Pro‑passkey side:
    • Passkeys are per‑site keypairs; private keys stay client‑side and aren’t reused elsewhere.
    • They’re resistant to phishing because credentials are bound to domains; browsers won’t use a credential on the wrong site.
    • They reduce password reuse and database breach impact.
  • Critics reply that:
    • If keys can be migrated, they can be stolen; any export mechanism becomes a high‑value phishing/malware target.
    • Relying on cloud providers to sync “device‑bound” keys already weakens the hardware‑bound story.

Backups, recovery & multi‑device usage

  • Big practical worry: losing a phone or account with synced passkeys could mean losing access to hundreds of services.
  • Suggested mitigations: multiple passkeys per account, backup hardware keys, or using a password manager as passkey provider.
  • Many sites only allow a single passkey and UX for adding multiple keys is inconsistent.
  • Some argue that traditional account recovery (email, backup codes) remains unchanged; others note this re‑introduces phishable fallbacks and undermines “passwordless” claims.

Hardware tokens vs software passkeys

  • Hardware tokens (e.g., FIDO keys) are praised for non‑extractable secrets, but criticized as impractical: limited storage for resident keys, easy to lose, and often not supported with multiple registrations.
  • Some want non‑resident keys and CA‑like models to avoid hardware storage limits.

Standards, governance & usability

  • Several complain the spec is complex and driven by large vendors with misaligned incentives (attestation, potential blocking of “too open” providers).
  • Others emphasize that designing something secure and usable “for billions” is intrinsically hard; passkeys already greatly help average users who are routinely phished via SMS/OTP.
  • Broad agreement that tooling, browser support, and UX (especially for non‑technical users) are still immature and uneven.