Summary
Logto's Account Center step-up logic accepted any existing verified WebAuthn registration record as proof of identity, as long as it belonged to the current user.
An attacker holding only an Account API bearer token could bypass proper step-up checks and manage MFA factors without re-proving possession of a password, identifier, or existing factor.
Affected context
This vulnerability appears in Logto deployments where the Account Center uses step-up checks to gate sensitive operations like MFA factor management.
The issue arises specifically when step-up authentication relies on generic "verified" records (including WebAuthn registration verifications) instead of requiring a fresh proof tied to the current high-risk action.
Environments that issue long-lived Account API bearer tokens and expose the Account Center routes are more exposed, as they provide more opportunities for misuse of stale or mis-scoped verification records.
Exploit preconditions
- Attack surface: Account Center and related routes that rely on step-up authentication for MFA and identity-sensitive operations.
- Required access: A valid Account API bearer token for the target user (e.g., from session hijacking, token theft, or unsafe storage).
- Trigger point: Creation and verification of a WebAuthn registration record for a new passkey using the bearer token, then reuse of that verification ID in the
logto-verification-idheader when calling Account Center routes. - Why exploitation is possible: The step-up logic trusted any verified record with
isVerified === truefor the user, instead of requiring a context-specific, recent, and possession-based verification event, effectively treating a registration verification as full identity proof.
Why this happens
The core issue is an improper authentication design where the step-up mechanism does not differentiate between verification types or contexts.
Rather than requiring a fresh, high-assurance check (such as password re-entry, OTP, or actual WebAuthn assertion), the system reused any existing verification record marked as isVerified === true for the user.
This maps to CWE-287: Improper Authentication, because the system incorrectly accepted a weaker verification artifact (a WebAuthn registration verification) as proof of full, current identity, allowing protected Account Center operations without a proper step-up challenge.
Potential impact
- An attacker with only an Account API bearer token could modify MFA factors for the victim's account.
- Attackers may add new passkeys or other factors they control, enabling persistent account access.
- Existing MFA protections can be weakened, removed, or replaced, reducing or eliminating strong authentication.
- In environments where Logto controls access to critical SaaS or AI applications, this may lead to full account takeover and downstream data exposure.
General mitigation guidance
Mitigation typically centers on tightening step-up authentication semantics and binding verifications to specific risk events.
Common defensive patterns include:
- Distinguishing clearly between registration-time verifications and authentication-time assertions, and never reusing registration verifications as step-up proof.
- Requiring a fresh, context-bound step-up (password re-entry, OTP, or WebAuthn assertion) immediately before sensitive actions such as MFA management or identifier changes.
- Scoping verification tokens by purpose and lifetime, ensuring they cannot be reused across different routes or operations.
- Designing APIs so that high-risk changes always involve a possession factor (device, key, or OTP) in addition to bearer tokens.
- Applying least-privilege principles to access tokens, limiting what Account API bearer tokens alone can perform without additional checks.
How teams detect issues like this
Detection often involves a mix of design review and runtime monitoring:
- Static and design analysis: Review authentication and step-up flows to ensure each sensitive route enforces the correct, context-specific verification type rather than generic flags like
isVerified === true. - Code review signals: Look for patterns where verification objects are reused across different flows (registration vs. authentication) or where step-up checks accept broad conditions instead of explicit factors.
- Automated checks: Security linters and policy-as-code tools can flag routes that perform sensitive operations without a preceding strong verification call or that rely solely on bearer tokens.
- Runtime logging and analytics: Monitor logs for high-risk actions (MFA changes, passkey additions) that occur without a corresponding recent login or step-up event, and alert on unusual patterns.

