What Is a Honeypot in Cybersecurity?
A jar of honey beside a laptop and a network cable on an office desk

What Is a Honeypot in Cybersecurity?

A honeypot in cybersecurity is a decoy: a fake server, service, account or file set up to look valuable to attackers while having no real business use. Because no legitimate user has a reason to touch it, any connection or login attempt is a strong sign of an intruder or a malicious bot. Defenders use honeypots to detect attacks early and to study how attackers work.

What does the term honeypot mean?

In security, a honeypot is bait: something that exists only to be found by attackers. NIST’s glossary defines it as a system, such as a web server, or a resource, such as a file on a server, designed to attract intruders “like honey is attractive to bears” (NIST CSRC glossary, 2022).

CISA’s 2026 guidance on cyber decoys describes honeypots as fake systems or services, often with deliberate weaknesses, that lure attackers in so defenders can watch them and keep them away from real assets (CISA, 2026). MITRE Engage, a framework for planning these operations, adds why the alerts are so reliable: a honeypot has no purpose except attracting and exposing attackers (MITRE Engage glossary, 2022).

Outside IT, the word also means a romantic trap in espionage or a kind of cryptocurrency scam. This article is about the honeypot cyber security teams use.

How does a honeypot work?

A honeypot works by looking like something worth attacking, sitting where no real user would go, and raising an alert the moment anyone touches it. Employees have no reason to sign in to a decoy, so contact with one is rarely innocent. A typical setup has five steps:

  1. Build the bait. Make a decoy that fits your network, such as a file server named for payroll or an admin account that looks forgotten.
  2. Place it where only an intruder would look. Put it on the internal network, along the path an attacker would follow after breaking in.
  3. Wire it to alerts. CISA notes that any decoy becomes a tripwire once it feeds your monitoring and alerting (CISA, 2026).
  4. Contain it. Isolate the decoy so an attacker who takes it over cannot reach real systems.
  5. Respond and learn. Treat an alert as a likely intrusion, follow your incident response plan, and adjust the decoys afterward.

Because decoy alerts are rare, they cut through the noise. CISA says decoys produce high-fidelity alerts and are especially useful against attackers who sign in with stolen credentials and use tools already on your systems, activity that signature-based tools often miss (CISA, 2026). One example from its guidance: route unused internal IP addresses to a dead end, so any scan of that range raises an alarm.

What are the types of honeypots?

The main types are low-interaction and high-interaction honeypots, research and production honeypots, honeynets, and honeytokens such as canary files. On websites, the word also names a hidden form field that catches spam bots.

Low-interaction and high-interaction honeypots

A low-interaction honeypot imitates a few services, such as a login prompt or an open port, with no real operating system behind them. It is cheap and fairly safe, but it captures only the first steps of an attack, and careful attackers can often tell it is fake. MITRE Engage’s glossary uses the word honeypot for exactly this kind of low-interaction, often low-fidelity decoy (MITRE Engage glossary, 2022).

A high-interaction honeypot runs real systems, so an attacker can sign in, run commands and move around while defenders watch. The information is far richer, but so are the cost and the risk: if containment fails, the attacker holds a real computer on your network. CISA rates honeypots medium to high in complexity and says they take a high level of security maturity to run (CISA, 2026).

Research and production honeypots

Research honeypots, run by universities, researchers and security vendors, usually sit on the open internet to study what attackers scan for. Production honeypots sit inside a company’s network to catch intruders who got past the perimeter. CISA’s guidance focuses on decoys inside your own environment rather than internet-facing honeypots (CISA, 2026), and that is the right frame for most businesses.

Honeynets

A honeynet is a network of honeypots made to look like a real environment, so defenders can watch an attacker move from system to system. Honeynets are mainly research tools, because convincing ones take real time and skill to build and contain.

Honeytokens and canary files

A honeytoken is decoy data rather than a decoy system. CISA defines honeytokens as data or objects with no legitimate business use, such as fake records, credentials or files, planted to detect unauthorized access or data theft (CISA, 2026). Common examples:

  • Sensitive-looking files that alert when opened, copied or changed
  • Fake records mixed in with real ones
  • Disabled or unused accounts, where any sign-in attempt is suspicious
  • Fake passwords or API keys left where an attacker would search
  • Decoy email addresses and web links that no real process uses

Canary files, sometimes called canary tokens, are honeytokens named for the canary in a coal mine. CISA rates honeytokens low in complexity and in the maturity needed to run them, which makes them the realistic starting point for smaller organizations (CISA, 2026). A stolen honeytoken can even show whose data was taken and how it leaked (MITRE Engage glossary, 2022). Spam traps apply the same idea to email: addresses kept to expose illegitimate senders, such as those who scrape or buy address lists (Spamhaus, 2022).

The hidden honeypot field on web forms

On websites, a honeypot is also a form field hidden from people. A human never fills it in, but a spam bot that completes every field gives itself away, and the form discards the entry. Real visitors never notice it, which makes it a low-friction partner or alternative to a CAPTCHA.

If you use one, hide the field from screen readers and browser autofill too, or real inquiries can be thrown away. Send yourself a test entry after any change to the form.

What do defenders learn from a honeypot?

A honeypot tells defenders that someone is probing or already inside, where they are, and how they work:

  • Early warning. A tripped decoy can be the first sign of an intruder using real credentials. CISA lists faster detection as a core benefit of decoys (CISA, 2026).
  • Location. The decoy that fired shows where the attacker is. Decoy files on an executive’s laptop point to that account.
  • Methods. What the attacker searched for and ran shows what to block next.
  • A test of your response. CISA recommends simulated attacks to confirm that decoy alerts fire and reach someone who acts on them (CISA, 2026).

What are the risks and limits of honeypots?

A honeypot only sees what touches it, and a poorly built one can give an attacker a new way in. Know the limits:

  • It detects; it does not prevent. A decoy raises an alarm but does nothing to stop an attack on your real systems.
  • It can be turned against you. A high-interaction honeypot that is not isolated can become a launch point for attacks on your network or on other companies.
  • Attackers can spot fakes. Experienced attackers look for services that respond oddly or systems with no real activity, and steer around them.
  • Someone has to watch it. An alert at 2 a.m. on a Saturday is worthless if nobody sees it until Monday. MITRE Engage warns that deploying a decoy is not the same as success and that monitoring is essential (MITRE Engage, 2022). Round-the-clock watching is the job of a security operations center (SOC).
  • It needs upkeep. Decoys must stay believable as your network changes, and IT testing can set them off. CISA recommends keeping the record of decoy locations outside the network it describes and limiting who knows the details (CISA, 2026).
  • It raises legal questions. MITRE Engage lists legal among the stakeholders to involve and stresses that engagement stays inside your own environment, never hacking back (MITRE Engage, 2022). CISA advises building basic detection decoys before observing attackers in depth, which takes isolated environments, mature logging and staff who can manage the legal risks (CISA, 2026).

Does a small business need a honeypot?

Most small businesses do not need to build and run a honeypot of their own. They get more from decoys built into managed detection and response, from logs that let someone investigate an alert, and from people who watch those alerts around the clock.

CISA calls its decoy guidance an introductory resource for small and medium-sized organizations, among others. It starts with simple tripwires and honeytokens, notes that these can run on tools you already own, such as endpoint detection and response, and saves deeper engagement for teams with mature monitoring and skilled staff (CISA, 2026).

Decoys you may already have

At OAC Technology, we run Huntress managed detection and response for clients’ computers, servers, Microsoft 365 accounts and firewalls, with Huntress analysts responding around the clock. Huntress includes Ransomware Canaries: decoy files hidden in folders that ransomware commonly targets, such as Documents. If one is changed, renamed or deleted, the Huntress agent alerts Huntress’s security operations center, which reviews it and issues an incident report if it finds signs of ransomware.

Huntress’s documentation says canaries are on by default, are not placed on network drives, and do not stop ransomware by themselves; automatic isolation of an infected computer needs its Managed Host Isolation feature turned on (Huntress, 2026). Huntress lists canary support for Windows 10 and 11 and Windows Server 2016 and later (Huntress, 2026). Pair them with the controls in our guide to preventing ransomware attacks.

What matters more than a honeypot

A decoy alert only helps if you can act on it. Before adding decoys, make sure you have:

  • Multifactor authentication on email, remote access and admin accounts
  • Regular patching of computers, servers and firewalls
  • Endpoint detection and response that someone watches day and night
  • Searchable logs: Microsoft 365’s standard audit log is on by default for eligible subscriptions and keeps records for 180 days (Microsoft Learn, 2026), and firewall and server logs need the same care
  • Tested backups

Our small business cybersecurity checklist lays out what a 10-, 25- and 50-person company needs, and our managed cybersecurity service includes vulnerability scanning and patch management. Once the basics are solid, a honeytoken or two is a low-cost extra tripwire, such as a disabled admin account that should never see a sign-in attempt.

Common questions

Can a honeypot be hacked?

Yes, and in a sense that is the point. The danger is a high-interaction honeypot that is not isolated: an attacker who takes it over can use it against your real systems or other organizations. Keep honeypots on separate network segments, watch them closely and rebuild them from a clean copy after each incident. Honeytokens carry far less of this risk.

What is a honeypot strategy?

A honeypot strategy is a plan for using decoys on purpose: which attacks you want to catch, what the decoys should look like, where attackers will find them, who gets the alerts and what happens next. MITRE Engage frames it as a repeating cycle of preparing, operating and understanding, and CISA’s 2026 guidance walks through that cycle with examples.

Is a honeypot the same as a sandbox?

No. A sandbox is an isolated space where security tools open suspicious files or links to see what they do before they reach users. A honeypot waits for attackers to come to it. Both rely on isolation, but a sandbox studies code you already have, while a honeypot reveals activity you did not know about.

Do honeypots replace antivirus or a firewall?

No. A honeypot only sees activity aimed at the decoy itself; it does not block malware, filter traffic or patch anything. Firewalls, endpoint protection, multifactor authentication and patching do the preventing, and decoys add an early warning when something gets past them. CISA presents decoys as a complement to Zero Trust security, not a substitute for it.

Will a honeypot attract attackers to my network?

Internal decoys and honeytokens do not advertise anything to the internet, so they draw no new attackers; they catch people who are already inside. An internet-facing honeypot will be scanned and attacked constantly, like any public address, so it must never share a network with your real systems. For most businesses, internal tripwires are the safer place to start.

Talk to us

If you want to know how an intruder would be caught at your business today, we can walk through your current detection, logging and alert handling with you and tell you plainly whether decoys would add anything. To talk it through, call us at 952-548-5558 or contact us. The first conversation is free, usually about half an hour, and there is no long-term contract.

Author picture

OAC Technology

Author

Related Articles