← Useful
17 September 20266 min read
News

The WordPress Administrator Trap: Grant Elevated Access Only When Needed

Permanent administrator access makes minor mistakes and stolen passwords unnecessarily serious. A role-based access model reduces risk without making the workday cumbersome.

On many business websites, editors, marketers, developers and external vendors work as administrators. This is convenient because no one is blocked by insufficient permissions. At the same time, each account gets access to far more than the person needs to perform their usual tasks.

The consequence is that a stolen password, an incorrect change or a compromised work computer can cause more extensive damage. An attacker who takes over an administrator account can, among other things, create new users, change settings, install code and attempt to hide their tracks.

The solution is not to make WordPress inaccessible to everyone except technical staff. The business needs an access model where routine work is carried out with limited permissions, while administrator access is granted in a controlled manner when a specific task requires it.

Tilganger bør vurderes ut fra faktiske arbeidsoppgaver.
Access should be assessed based on actual job responsibilities.

Start with the tasks, not the job titles

Access management is often based on job titles. A marketing manager becomes an administrator because they are responsible for the website, while an agency gets full access because it provides technical services. Neither necessarily says anything about which actions they actually need to perform.

Instead, create a simple overview of the tasks being carried out:

  • write and edit content
  • publish pages and posts
  • process forms, orders or products
  • change menus and other website settings
  • update WordPress, themes and plugins
  • install or remove components
  • create users and change permissions
  • investigate errors, logs and security alerts

Then assign each person or function the tasks they need. The goal is for the everyday account to have permissions only for routine work. Higher access should be a deliberate exception.

Divide access into four practical levels

WordPress has built-in roles, but they do not always match how work is divided in a business. Plugins and e-commerce solutions may also add their own permissions. What matters, therefore, is not just the role name, but what the account can actually do.

1. Content user

This level is suitable for people who write or maintain text and images. They normally do not need to install plugins, manage users or change technical settings. If content must be reviewed before publication, the permission to publish can also be restricted.

Sterk autentisering beskytter kontoer med utvidede rettigheter.
Strong authentication protects accounts with elevated permissions.

2. Subject matter or publishing manager

This user can approve and publish content, tidy up the media library and manage relevant content features. Access should still not include installing code, changing security settings or creating administrators.

3. Technical operations user

The operations user is used for scheduled updates, troubleshooting and technical changes. The account should not be the same one a developer or consultant uses for routine content work. When a technical task is complete, the account should be logged out and not used until the next agreed need.

4. Emergency administrator

This is an account for situations where normal administration does not work, for example if the authentication solution fails or other administrator accounts become locked. The login details must be stored securely, and all use should trigger a review and a new password. The account must not be a shortcut in everyday work.

Strong authentication must apply to the right accounts

Two-factor authentication is particularly important for accounts that can publish, change settings or access personal data. But an additional login step does not solve everything. The business must also control how accounts are created, used and recovered.

  • Use personal accounts: Shared administrator users make it difficult to know who performed an action.
  • Avoid reused passwords: Each account must have a unique, long password managed in a suitable password solution.
  • Protect recovery: The email account that can reset the WordPress password must be at least as well secured as the website itself.
  • Control backup codes: Recovery codes must be stored securely and must not be left in an open folder or email thread.
  • Record technical keys: Integrations, application passwords and other machine access must have a named owner and a clear purpose.

An account with no known owner should be treated as an exception. The same applies to old integrations that no one can any longer explain the need for.

Sikkerhetsvarsler må ha en tydelig ansvarlig mottaker.
Security alerts must have a clearly designated recipient.

Turn admin access into a controlled work session

When someone needs elevated access, there should be a short, predictable routine. It need not be bureaucratic, but it must make clear what is going to happen.

  1. Describe the task and why admin access is necessary.
  2. Check that a recent backup exists and that there is a known way back.
  3. Grant access to a named individual, not a shared user.
  4. Perform the change within the agreed time frame.
  5. Test the most important functions afterward.
  6. Document what was changed.
  7. Remove or reduce access when the work is complete.

A simple example is installing a new form feature. The provider does not receive a permanent administrator role because the solution may need adjustments later. Access is granted for installation and verification, then reduced when the task is complete.

Connect updates and vulnerabilities to the access model

Updates are safer when responsibilities and permissions are aligned. It should be clear who assesses security alerts, who can approve an urgent change, and who actually carries out the update.

Not all updates require the same process. A routine maintenance update can be included in a planned maintenance window. A known vulnerability affecting the website's active components may require faster action. The team must then be able to answer three questions:

  • Does the website use the affected component or function?
  • Is a secure update or temporary risk-mitigation measure available?
  • Who has the authority and access to carry out the change now?

If everyone is listed as an administrator, responsibility is not necessarily clearer. Often the opposite is true: Everyone can do something, but no one knows who is supposed to do it.

Backups must be protected from the same administrators

An administrator account that can change the website should not automatically be able to delete all backups. If both the production environment and the backup are managed through the same account or control panel, a single compromise can affect both.

Backups should therefore be stored separately from the website, with their own access controls. It must also be clear who can initiate a restoration and how that action is controlled. Restoring should be practiced, but the day-to-day administrator does not necessarily need full control over the entire backup history.

Monitor high-impact actions

Monitoring should prioritize events that can change the security level. Large volumes of technical logs are of little value if no one knows which signals require action.

Pay particular attention to:

  • creation of new administrators
  • changes to roles and permissions
  • unusual or repeated failed logins
  • installation, activation, and deletion of extensions
  • changes to authentication and security settings
  • use of the emergency account
  • unexpected changes outside agreed work windows

Alerts must have a recipient who understands what the event means and has the authority to respond. An alert that simply accumulates in an overflowing inbox is not real monitoring.

Implement the model in 30 days

The company does not need to rebuild its entire operation at once. A controlled cleanup can be carried out in four steps.

  1. Week 1: Export or record all users, roles, technical accounts, and integrations. Identify accounts without a clear owner.
  2. Week 2: Map the tasks each user actually performs. Reduce permissions where administrator-level access is not necessary.
  3. Week 3: Implement strong authentication, secure recovery methods, and establish a separate emergency account.
  4. Week 4: Agree on the procedure for temporary elevated access, relevant alerts, and a regular review of users and permissions.

The review should be repeated regularly and after major changes to the organization, its providers, or the website. Check not only who has access, but also which new permissions extensions and integrations have introduced.

Good access management should have little impact on everyday work

A secure model works when the editor can publish without technical access, the developer can carry out an agreed change without permanent administrator status, and management knows who can do what. It should not create unnecessary obstacles, but limit the impact when an account, a machine, or a work process fails.

The most important first step is simple: Find all permanent administrators and ask which specific, recurring task requires this level of access. If the answer is unclear, access should be reduced.