Explore our new Cybersecurity Terms Glossary, explained by our experts.
Cybersecurity Terms Glossary
ABAC decides each request by evaluating attributes of the user, the resource, the action, and the context, rather than by checking which role the user happens to hold.

What Is Attribute-Based Access Control?
Attribute-based access control answers a different question from the one roles answer. Instead of asking “what role does this person hold?” it asks “given everything I know about this user, this record, this action, and this moment, should it be allowed?” The decision is computed per request from attributes rather than looked up in a membership list.
A policy reads like a sentence. A clinician may open a patient record if the patient is assigned to their ward, the request arrives during a rostered shift, and the device is managed. None of that fits in a group name, which is precisely the point.
NIST Special Publication 800-162 sets out this model in detail, along with the architectural split between the policy decision point that evaluates rules and the policy enforcement point that applies the answer. Keeping those separate is what lets you change a rule without redeploying an application.
| RBAC | ABAC | |
|---|---|---|
| Decision basis | Role membership | Attributes evaluated at request time |
| Granularity | Coarse; per function | Fine; per record and context |
| Change cost | New role, new assignment | Edit a policy |
| Auditability | Easy to list who holds what | Requires simulation and decision logs |
| Main failure mode | Role explosion | Bad attribute data |
The classic argument for ABAC is role explosion: when every combination of region, product line, and seniority needs its own role, a mid-sized organisation ends up maintaining several thousand of them. Most mature deployments do not choose between the two. They use role-based access control for the broad grant and attributes to narrow it, which keeps the “who has access to what” question answerable while still allowing contextual rules.
In practice ABAC shows up in three places. Cloud platforms use it natively: AWS IAM policy conditions on resource tags and principal tags, Azure attribute-based access control on storage blobs, and Google Cloud IAM conditions all evaluate attributes at request time. Application-layer policy engines such as Open Policy Agent, Cedar, or a commercial externalised authorisation service let services ask a shared decision endpoint instead of each one hard-coding rules. And conditional access, covered in our entry on conditional access policies, is ABAC applied at the point of authentication.
Each of those needs attributes delivered to the decision point. That usually means claims in a JSON Web Token, which introduces its own constraint: a token minted ten minutes ago carries attributes as they were ten minutes ago. For attributes that change quickly—device compliance, risk score, employment status—either keep token lifetimes short or have the decision point fetch live values rather than trusting the token.
ABAC is only as good as the HR and asset data behind it, and that data is usually worse than anyone believes. If the directory says 40% of staff are in “Head Office” because nobody updated the field after the office moved, a location-based policy is decoration. If contractor end dates are blank, a policy that revokes on expiry revokes nothing.
Before writing policies, profile the attributes you intend to rely on: completeness, freshness, who owns the authoritative source, and what happens when a value is missing. Decide the default for absent attributes explicitly—deny is the safe answer, and it will surface data gaps quickly. Then test policies in a simulation mode that logs the decision without enforcing it, long enough to see the edge cases that nobody described during design.
Finally, log the decision with its inputs. The great weakness of ABAC against auditors is that you cannot produce a list of who can see a given record without evaluating policy against every user. Decision logs that record the attribute values used, not just the allow or deny outcome, are what make an ABAC deployment defensible under ISO 27001 or UK GDPR accountability questions about who accessed personal data and on what basis.
At Privia Security, we review authorisation design and test it against real abuse cases in our web application security services.
Rarely. The pattern that works for most organisations is a hybrid: roles grant the broad entitlement, attributes narrow it to the specific records and contexts the person should see. Pure ABAC makes the question “who can access this?” expensive to answer, which auditors and incident responders both need. Pure RBAC leads to thousands of near-duplicate roles.
It is the component that evaluates the policy and returns allow or deny, separate from the enforcement point in the application that acts on the answer. Separating them means a rule change takes effect everywhere without redeploying code, and it gives you one place to log every decision. Open Policy Agent and Cedar are common implementations.
Usually as claims inside an identity or access token issued by the identity provider, or by the decision point querying an authoritative source directly. Tokens are faster but carry a snapshot: anything that changes quickly, such as device compliance or employment status, may be stale. Short token lifetimes or live lookups for volatile attributes resolve that trade-off.
Decide that explicitly rather than letting the engine choose. Defaulting to deny is the safe option and has the useful side effect of exposing data quality problems immediately. Defaulting to allow turns every blank field in the HR system into an open door. Either way, log missing attributes separately so the owners of the source data can fix them.
You May Be Interested In These
We use a few cookies that are needed to make this website work, and they are always on. We would also like to set analytics cookies, to see which pages are useful; functional cookies, so that our live chat works; and advertising cookies, to measure our campaigns. We will not set any of those until you say yes, and you can change your mind at any time from the Cookie Settings link in the footer.
Choose the cookies you are happy for us to use. Strictly necessary cookies are always on; everything else is off until you turn it on.
These cookies keep the website working: they hold your security checks, your form submissions and the choice you make here. The site cannot run without them, so they are always on.
These let our live chat open and remember your conversation, and let the newsletter sign-up work. Switch them off and the chat window will not appear.
These tell us which pages are read and how people find us, so that we can improve the site. The reports are about groups of visitors, never about you personally.
These let us measure our advertising and show you our adverts on other websites, such as LinkedIn, Google and Meta. Switch them off and you will still see adverts, but they will be less relevant.