Explore our new Cybersecurity Terms Glossary, explained by our experts.

View the Glossary View the Glossary
PRIVIA

Cybersecurity Terms Glossary

What Are Canary Accounts?

A canary account is a convincing but unused identity planted in a directory. Nobody has a legitimate reason to touch it, so any authentication attempt against it is a detection with almost no false positives.

Illustration for the cybersecurity term Canary Accounts
Home / Blog / Cybersecurity Terms / What Are Canary Accounts?
17 November 2026

What Are Canary Accounts?

What Are Canary Accounts?

A canary account is a deliberately planted identity that looks valuable and is never used. Because no employee, script, or service has any reason to authenticate as it, an attempt to do so is a near-certain sign of unauthorised activity. That property makes canaries unusually precise in a discipline where most detections are compromises between sensitivity and noise.

Why They Work So Well

Attackers who gain a foothold in an Active Directory environment follow a predictable sequence: enumerate accounts, look for ones with elevated privileges or service principal names, and attempt to obtain their credentials. They cannot tell which entries in the directory are real. A well-constructed decoy named for a departed finance director, or a service account registered with an SPN that looks like a legacy SQL connection, is exactly the kind of target that enumeration tooling ranks highly.

Compare that with a conventional detection such as alerting on unusual logon times, which depends on a baseline, needs tuning per department, and generates false positives every time somebody works late. The canary requires no baseline at all. The alert condition is simply: this identity was touched.

Varieties Worth Deploying

Building One That Survives Inspection

Most canaries fail because they look new. A good decoy has a creation date consistent with its backstory, a password last-set date in the past, a populated description, department, and manager field, and group memberships that match the pattern of comparable real accounts. If every real service account in your directory has a description referencing a change ticket and the canary does not, an attacker who has already read the directory will notice.

Equally important is making sure nothing legitimate touches it. Vulnerability scanners, asset discovery tools, compliance scripts, and identity governance reviews all enumerate directories, and the first week of a canary deployment is usually spent identifying which internal tool is generating the alerts. Exclude those sources explicitly rather than widening the alert conditions.

Document the canaries somewhere that attackers will not reach but responders will. Microsoft Defender for Identity supports tagging accounts as honeytokens, which raises the severity of any activity involving them, and similar labelling exists in most SIEM platforms through a simple lookup list.

Canary Accounts and the Wider Family

Accounts are one instance of a broader approach described under deception technology, which also covers fake file shares, decoy hosts, and canary tokens embedded in documents. The accounts are usually the right place to start, because they cost nothing, need no new infrastructure, and detect the stage of an intrusion where defenders most need warning.

What they do not do is cover initial access or anything that happens before an attacker reaches the directory, and a canary that fires is already evidence that something earlier was missed. Treat them as a high-confidence tripwire layered on top of monitoring, not a replacement for it. At Privia Security, we deploy canary identities as part of our SOC maturation services, usually in the first fortnight, because the return on effort is immediate.

FAQ – Frequently Asked Questions

How Many Canary Accounts Should an Organisation Create?

A handful placed thoughtfully beats dozens scattered at random. Three to five in Active Directory covering different attacker interests, one or two in each cloud tenant, and credential bait on jump hosts and administrative workstations is a reasonable starting position. Too many become difficult to maintain convincingly, and inconsistency between them is itself a signal that an attentive attacker can use.

Can an Attacker Tell a Canary Account From a Real One?

A careless one, yes. Accounts created on the same day with sequential names, empty description fields, recent password-set dates, and no group memberships stand out immediately. Build decoys to match the metadata pattern of your genuine accounts, age them before relying on them, and refresh attributes occasionally so they do not become the only accounts in the directory that never change.

What Should Happen When a Canary Account Alert Fires?

Treat it as high severity by default and investigate immediately. Because legitimate interaction is close to zero, the base rate strongly favours real malicious activity, and the usual priority rules should be bypassed. Identify the source host and account, preserve its volatile data, and scope outwards from there. Record and exclude any internal scanning tool that proves to be the cause.

Are Canary Accounts Legal to Use?

Yes. They are decoys inside your own estate, monitored like any other system, and they do not interfere with anyone’s data or systems. Keep the monitoring consistent with your privacy notice and employee monitoring policy, since alerts will occasionally involve staff activity, and avoid any design that would attack or interfere with a third party’s systems, which would raise entirely different legal questions.

Related Terms

You May Be Interested In These

What Is the Chain of Custody?

Cybersecurity Terms

17 November 2026

What Is the Chain of Custody?

Read More