
A honeytoken is a piece of data that has no legitimate use, planted where an intruder is likely to find it. It might be a fake password in a configuration file, a database row describing a customer who does not exist, or a document that quietly reports when it is opened. Nobody should ever touch it. If someone does, you have learned something important.
Why they work
Attackers who reach a network need to find credentials, data and paths to more systems. They search file shares, read scripts, dump password stores and query directories. Honeytokens sit right on that path. Unlike a full decoy server, they are cheap, quick to deploy and almost invisible to normal users, so the alerts they produce are rare and meaningful.
Common types
| Honeytoken | Where to place it | What triggers the alert |
|---|---|---|
| Fake cloud access key | Developer laptops, code repositories, CI configuration | Any API call made with the key |
| Decoy directory account | Active Directory, with an attractive name and description | Login attempts or ticket requests for the account |
| Canary database record | Customer or player tables | The record's email address or phone number receiving contact |
| Beacon document | Finance or HR shares | The file being opened and calling home |
| DNS token | Scripts, bookmarks, internal wikis | A lookup of a unique hostname |
Free and commercial services can generate many of these and handle the alerting for you. You can also build simple ones yourself, for example a disabled account whose login attempts are forwarded straight to your alerting pipeline.
Placement and hygiene
- Make tokens look plausible. A file called passwords.txt in an obvious place may catch a careless intruder; a realistic service account in a real deployment script catches a careful one.
- Document every token in a private register, so your own team does not trigger them during audits.
- Exclude tokens from automated scanners you control, or you will train analysts to ignore the alerts.
- Refresh them when systems change. A token in a server that was retired years ago protects nothing.
Beyond the corporate network
Canary records are especially useful wherever personal data has value on the black market. A betting operator can place a few synthetic player profiles in its customer database; if those email addresses ever receive spam or phishing, the data has leaked somewhere, even if no alarm went off inside. The same trick works for loyalty programmes, ticketing systems and any list that criminals like to sell.
Honeytokens will not stop an attack on their own. What they do is turn the attacker's curiosity into an early warning, at very low cost. For the bigger picture of where they fit, see our overview of decoy networks.