Role and access mistakes that expose other people's data in a cabinet

The most expensive access mistakes look mundane

When a business thinks about security risk it often imagines a sophisticated external attack. In customer cabinets and B2B systems the painful problems are usually simpler: a user sees a document that is not theirs, a manager gets access that is too wide, an API does not check who owns the object it returns.

Where access control most often breaks

• The role is checked but the right to a specific object is not — the classic object-level access gap.
• One role serves too many different scenarios, because the system was grown to "not get in the way".
• The API trusts an object identifier from the client instead of deriving ownership from the session.
• Admin screens and background jobs bypass the same checks the user interface enforces.
• Export, bulk actions and service APIs open workarounds around the main logic.

Why a cabinet is especially sensitive

Because that is where documents, statuses, orders, financial data, operation history and the user's role in the process live. The more a cabinet resembles a real working tool, the higher the cost of one access mistake. Here security is not a separate checkbox — it is part of the product logic.

What to verify besides the roles themselves

• Which objects a user should see and change within their role.
• How the server checks ownership of each object: order, document, client, contract, ticket.
• Whether background integrations, exports, bulk actions and service APIs open bypass routes.
• Whether critical actions are logged well enough to investigate an incident.

What a mature outcome looks like

Not an abstract "we fixed the permissions", but a managed access model: a role matrix, object-level checks, clear contour boundaries, logging of critical actions and a remediation plan that genuinely reduces the risk of a leak or unauthorised operations.