Summary
A GraphQL authorization flaw in Craft CMS can expose personal information belonging to users who created drafts or revisions. A client with only draft or revision read access—and, in some configurations, an unauthenticated public GraphQL client—may retrieve creators' email addresses, usernames, full names, and postal addresses.
Affected context
The issue occurs in Craft CMS deployments that expose the GraphQL API and permit access to draft or revision data. The draftCreator and revisionCreator fields rely on element-level scopes rather than the user-data authorization required for user information, and the returned User element does not consistently enforce authorization for sensitive fields. Risk is higher when the public GraphQL schema is enabled with these scopes.
Exploit preconditions
- Attack surface: Craft CMS GraphQL API
- Required access: A client with
elements.drafts:readorelements.revisions:read; unauthenticated access may be possible when the public schema is configured with those scopes - Trigger point: Querying draft or revision creator fields through GraphQL
- Why exploitation is possible: Field authorization checks validate draft or revision access but do not require the user-data scope or consistently restrict sensitive User fields
Why this happens
The vulnerability results from an authorization boundary mismatch. GraphQL fields exposing draft and revision creators are gated by element scopes instead of the user-data permission enforced for user queries. The resolver then returns a raw User element whose sensitive fields lack appropriate per-field authorization, mapping to CWE-200: Exposure of Sensitive Information to an Unauthorized Actor.
Potential impact
- Attackers may harvest email addresses and usernames
- Personal names and postal addresses could be exposed
- Privacy obligations and data-protection requirements may be affected
- Exposed information could support phishing, impersonation, or targeted attacks
- Information about site editors and administrators may be disclosed
General mitigation guidance
Mitigation generally focuses on aligning GraphQL field authorization with the sensitivity of the returned data.
Common defensive patterns include:
- Requiring explicit user-data permissions before resolving creator identity fields
- Applying per-field authorization to email, username, full name, and address data
- Returning the minimum identity information needed by each client
- Keeping public GraphQL schemas disabled unless their exposure is intentional and reviewed
- Restricting draft and revision scopes according to least privilege
- Treating personal data as a separate trust boundary from content metadata
- Auditing existing API clients and rotating access where unnecessary exposure may have occurred
How teams detect issues like this
Organizations can identify similar issues through:
- Static analysis of GraphQL resolvers for authorization checks based only on parent-object permissions
- Code review for raw model or entity returns from content-oriented fields
- Tests verifying that each sensitive field enforces its own authorization policy
- GraphQL schema reviews that distinguish public, content, and user-data scopes
- API monitoring for broad enumeration of creator fields and repeated requests across drafts or revisions
- Policy enforcement that prevents public schemas from exposing personal-data fields without explicit approval

