Read a sample: 1. What an Access Review Accomplishes
A small company uses an online billing application to send invoices, receive payments, and maintain customer records. One morning, the operations manager asks a reasonable question: “Who can change our bank account details?”
The IT administrator knows who currently works in billing, but the application has been in use for several years. A former employee helped set it up. An outside bookkeeper was given access during a busy season. A few people have administrator roles because that was the quickest way to get them working. No one is certain which roles can change payment settings.
That question is the beginning of an access review. It is not yet a review by itself. To answer it responsibly, the team must find out what access exists, determine which permissions matter, ask knowledgeable people whether the access is still appropriate, and record what happens next.
An access review is a structured examination of who or what can use a system and what each account is allowed to do. The purpose is to compare existing access with current, authorized business needs. The review may confirm that access is appropriate, identify a change, or reveal a question that requires further investigation.
For a small team, the process does not need to begin with specialized software or an elaborate committee. It does need a defined scope, reliable enough information to make decisions, someone accountable for those decisions, and follow-through when access must change.
Access is usually granted for a reason. A new employee needs an account, a manager needs a report, or a vendor needs to investigate a problem. At the time, the request may be sensible and properly approved. The difficulty is that the reason can change while the access remains.
People move between responsibilities. Temporary projects become permanent, or they end without a clear handoff. Applications acquire new features, and a role that once seemed narrow may now include additional capabilities. A vendor engagement expires, but an account is left active. An administrator grants broad access to solve an urgent problem and intends to narrow it later.
None of these situations requires bad intent. Access can drift simply because businesses change faster than account records are revisited.
A periodic review creates a deliberate opportunity to ask whether the original decision still makes sense. It also gives the organization a record of the answer. Instead of relying on someone’s memory that an account was “probably needed,” the team can identify the purpose, the person responsible for the decision, and any action required.
The review is not limited to finding accounts that should disappear. It can also expose unclear ownership, poorly understood roles, duplicate accounts, or a gap between what a manager believes an employee can do and what the application actually permits. Those findings may require separate work, but they are valuable because they make hidden assumptions visible.
Several terms appear throughout this book. They are related, but they answer different questions.
An identity represents a person or nonhuman actor, such as an automated process. An account is an identity’s means of access within a particular system. One person may have more than one account, and some accounts may not belong to an individual person at all.
Authentication is how a system checks that someone or something attempting to sign in is the expected account holder. Authorization concerns what that authenticated account is permitted to do. Signing in successfully does not, by itself, explain whether the account can read customer records, issue refunds, change configuration, or administer other users.
An access review focuses on authorization and the continuing need for it. The reviewer asks questions such as:
- Is this account still needed, and can its identity or purpose be established?
- What can the account actually do, including through groups or assigned roles?
- Is that level of access appropriate for its current business purpose?
- Who has the authority and knowledge to confirm the decision?
These questions apply to people and to nonhuman accounts. An automation account may not have a manager in the ordinary sense, but someone should still understand what it does and be accountable for its use.
The distinction between a role’s name and its capabilities is especially important. A role called “manager” might allow only reporting in one application and extensive configuration changes in another. A reviewer who approves access based solely on a reassuring role name may miss significant permissions. When the capabilities are unclear, the correct decision may be to investigate rather than guess.
Least privilege means giving an account the access needed for its authorized purpose, without unnecessary additional privileges. It is a standard for making a decision, not a demand that every account have the fewest permissions imaginable.
A billing specialist who prepares invoices may need to view customer details and create invoices. Whether that person also needs to change the bank account that receives payments is a separate question. The answer depends on the work, the available roles, and the controls the business has chosen.
Applying least privilege requires judgment because application permissions rarely fit every job perfectly. A vendor may package several capabilities into one role, even though the employee needs only some of them. A team may need temporary elevated access to complete a migration. An automated process may require a permission that appears broad because the application offers no narrower option.
In these cases, the reviewer should understand and document the…