---
id: CP-039
slug: dictionary-attack-mutation-rules-passwords
lang: en
title: Dictionary attacks and mutation rules
description: How an attacker tries password lists, then their variants (capital letter, digit, @), and how to protect yourself.
canonical_url: https://startbrain.io/cyberpedia/en/articles/dictionary-attack-mutation-rules-passwords/
markdown_url: https://startbrain.io/cyberpedia/en/articles/dictionary-attack-mutation-rules-passwords.md
publisher: Brain (https://startbrain.io)
license: CC BY 4.0 (https://creativecommons.org/licenses/by/4.0/)
cite_as: Brain, "Dictionary attacks and mutation rules", Cyberpedia, 2026. https://startbrain.io/cyberpedia/en/articles/dictionary-attack-mutation-rules-passwords/ (CC BY 4.0)
wing: 1
theme: master_passwords
reflex: cr-011
schema: G3
updated: "2026-11-14"
translations:
  en: https://startbrain.io/cyberpedia/en/articles/dictionary-attack-mutation-rules-passwords/
  fr: https://startbrain.io/cyberpedia/fr/articles/attaque-dictionnaire-regles-mutation-mots-de-passe/
  de: https://startbrain.io/cyberpedia/de/articles/woerterbuchangriff-mutationsregeln-passwoerter/
  es: https://startbrain.io/cyberpedia/es/articles/ataque-diccionario-reglas-mutacion-contrasenas/
  it: https://startbrain.io/cyberpedia/it/articles/attacco-dizionario-regole-mutazione-password/
  nl: https://startbrain.io/cyberpedia/nl/articles/woordenboekaanval-mutatieregels-wachtwoorden/
  pl: https://startbrain.io/cyberpedia/pl/articles/atak-slownikowy-reguly-mutacji-hasel/
  pt: https://startbrain.io/cyberpedia/pt/articles/ataque-dicionario-regras-mutacao-palavras-passe/
  ja: https://startbrain.io/cyberpedia/ja/articles/jisho-kogeki-mutation-rule-password/
sources:
  - "n": 1
    title: Blocking Brute Force Attacks
    publisher: OWASP
    url: https://community.owasp.org/controls/Blocking_Brute_Force_Attacks
    accessed: "2026-11-14"
  - "n": 2
    title: "Mots de passe : recommandations pour maîtriser sa sécurité"
    publisher: CNIL
    url: https://www.cnil.fr/fr/mots-de-passe-recommandations-pour-maitriser-sa-securite
    published: "2022-10-14"
    accessed: "2026-11-14"
  - "n": 3
    title: What Is a Brute-Force Attack?
    publisher: Proofpoint
    url: https://www.proofpoint.com/us/threat-reference/brute-force-attack
    accessed: "2026-11-14"
  - "n": 4
    title: "Password policy: updating your approach"
    publisher: NCSC (UK)
    url: https://www.ncsc.gov.uk/collection/passwords/updating-your-approach
    accessed: "2026-11-14"
  - "n": 5
    title: Brute Force Attack
    publisher: OWASP
    url: https://community.owasp.org/attacks/Brute_force_attack
    accessed: "2026-11-14"
  - "n": 6
    title: NIST Special Publication 800-63B
    publisher: NIST
    url: https://pages.nist.gov/800-63-4/sp800-63b.html
    accessed: "2026-11-14"
  - "n": 7
    title: Brute Force, Technique T1110 - Enterprise | MITRE ATT&CK®
    publisher: MITRE ATT&CK
    url: https://attack.mitre.org/techniques/T1110/
    accessed: "2026-11-14"
  - "n": 8
    title: "Brute Force: Password Cracking, Sub-technique T1110.002 - Enterprise | MITRE ATT&CK®"
    publisher: MITRE ATT&CK
    url: https://attack.mitre.org/techniques/T1110/002/
    accessed: "2026-11-14"
  - "n": 9
    title: "Brute Force: Password Guessing, Sub-technique T1110.001 - Enterprise | MITRE ATT&CK®"
    publisher: MITRE ATT&CK
    url: https://attack.mitre.org/techniques/T1110/001/
    accessed: "2026-11-14"
  - "n": 10
    title: Russian State-Sponsored Cyber Actors Gain Network Access by Exploiting Default Multifactor Authentication Protocols and “PrintNightmare” Vulnerability
    publisher: CISA
    url: https://www.cisa.gov/news-events/cybersecurity-advisories/aa22-074a
    published: "2022-03-15"
    accessed: "2026-11-14"
---

# Dictionary attacks and mutation rules

To speed up a brute-force attack, an attacker can start with dictionary words and slightly modified passwords, because most people choose those rather than a completely random password. OWASP files these attacks under the names dictionary attacks or hybrid brute-force attacks [1].

A password that looks complicated but follows a common pattern is among the variants that get tested. Starting from the word "kangourou", the CNIL gives for example "k4ng0urou", "kangourou01" and "KaNgOuRoU" among the derivations that attackers build and try [2].

## In 30 seconds

- Where it happens: on exposed login pages (web, SSH, RDP, VPN) and offline, against stolen password databases stored as hashes [3].
- What the attacker wants: to recover your password and log in as you [4].
- The reflex: a long passphrase, unique to each service (the CNIL cites one of at least 7 words), kept in a password manager, with two-factor authentication wherever it exists [2][4].
- If it's already done: change that password as soon as you know or suspect it is compromised, along with the password of any other service where you used it [4][2].

## Straight answers

### What is a dictionary attack?

The attacker prepares values in advance, sends them to a server and analyzes the response. To go faster, they can use a dictionary attack, with or without mutations [5]. The dictionary contains obvious choices such as "password123" and "qwerty", but also millions of passwords from earlier breaches [3].

### What are mutation rules?

They are the transformations applied to each word in the list: adding digits or special characters, or replacing characters following a planned sequence. Proofpoint gives the example of "password", "password1", "password!" and "p@ssword" [3]. The NCSC cites replacing the letter "o" with a zero as a predictable strategy [4].

According to OWASP, widely available tools rely on wordlists and rulesets that guess passwords "intelligently" [1]. According to Proofpoint, Hashcat takes advantage of the power of graphics cards and offers dictionary, combinator, mask and hybrid modes, while John the Ripper supports many hash types [3].

### Why does a password that looks complicated fall quickly?

Because apparent complexity does not measure randomness. According to the CNIL, any construction rule reduces the number of possible choices, and picking a word from a language very strongly limits the combinations in practice [2]. The NCSC adds that, to meet a complexity requirement, many users follow predictable patterns that attackers know and use to optimize their attacks [4]. According to Proofpoint, a long password that appears in leaked databases or follows an easy-to-guess pattern falls much faster than pure brute force would suggest [3].

### Is a long passphrase enough?

It helps a great deal. The CNIL sets a minimum of 80 bits of entropy for a password with no complementary measure, which a passphrase of at least 7 words satisfies, for example [2]. The NCSC recommends the "three random words" technique to get memorable passphrases [4]. NIST requires at least 15 characters for a password used alone, and 8 if it is used only within multifactor authentication [6].

### Should you change your password regularly?

Not without a reason. The CNIL has dropped the renewal requirement for standard accounts and keeps it for privileged accounts. It notes that most users reuse a slightly modified version of the old password, for example by adding a digit at the end [2]. NIST forbids requiring a periodic change, but requires a change if there is evidence of compromise [6].

### What can an employee and a company do?

For the employee:

- choose a long passphrase rather than a "transformed" word, avoiding personal information such as names, dates or sports teams [4];
- do not reuse a password between work and home, and use a password manager with its built-in generator [4];
- turn on two-factor authentication: even if an attacker guesses the password, the second factor is missing [3].

For the company:

- refuse overly common passwords with a deny list, either published or tailored to the organization [4];
- limit attempts, with a growing delay between tries (preferred by the NCSC) or a lockout after 5 to 10 attempts [4];
- require two-factor authentication for important accounts and for services exposed to the Internet [4].

## Inside the attacker's head

According to MITRE ATT&CK, the attacker turns to brute force when they do not know the passwords or when they have obtained hashes. They can query a service that checks credentials, or work offline on data already stolen [7]. Brute force comes into play at different moments of an intrusion, for example against valid accounts spotted after other actions inside the victim's network [7].

When they hold hashes, they try to recover the plaintext password. MITRE specifies that this cracking is usually done on systems the attacker controls, outside the target network, and that the password obtained is then used to log in to the account's resources [8]. Proofpoint stresses that this lets them bypass account lockouts without alerting the organization, using specialized hardware able to test billions of combinations per second [3].

Online, they can guess from a list of common passwords, whether or not they take the target's complexity policy into account. MITRE considers this route risky: it can cause many failed logins and lockouts [9]. The targeted services are administration services (SSH, RDP, LDAP, Kerberos), but also cloud applications and email such as Office 365 [9]. MITRE gives as examples the Pony malware, which used a small dictionary of common passwords against a list of local accounts, and Emotet, which used a hard-coded password list [9].

## What you see on screen

Most of the time, nothing. OWASP notes that an account lockout is ineffective against a slow attack that tries only a few passwords per hour [1].

Some signs can appear:

- an account locked for no reason, which the NCSC lists among the signals to watch [4];
- a two-factor code received without having requested it: with some methods (SMS or email), the NCSC sees this as a warning of compromise and asks that it be reported [4];
- on the defenders' side, a series of failed logins followed by a success on the same account, or attempts outside working hours [7][3].

Quick reporting to the IT department is what the NCSC encourages [4]. Without two-factor authentication, an attacker who finds the right password immediately gets everything the account can reach [3].

## Under the hood

A human types a few passwords per minute, a computer can process hundreds or thousands, and a single computer can test thousands per second [3]. Attackers use this at scale: OWASP describes tools that route each request through a list of proxy servers, so that every attempt seems to come from a different address. With 1,000 proxies, an attacker can try 2,000 or 3,000 passwords without being blocked [1].

Account lockout therefore has limits. OWASP lists several: an attacker can lock out legitimate accounts in bulk (denial of service), can try a single password across a large list of usernames, and administration accounts often escape the lockout [1]. The NCSC prefers a growing delay, which limits attempts while still giving the user several chances [4].

On the storage side, NIST requires passwords to be salted and hashed with a suitable scheme, whose cost makes each attempt more expensive for whoever stole the file [6]. The CNIL cites scrypt and Argon2 [2]. Salt makes it harder to use precomputed tables (rainbow tables) [3].

Finally, NIST asks that any new password be compared against a list of common, expected or compromised values: passwords from breaches, dictionary words, and context-specific words such as the name of the service or of the user. The password must be compared in full, and the user must be guided so as not to settle for a trivial modification of a rejected word [6].

## Case files

### May 2021: an inactive account guessed, an NGO infiltrated

According to advisory AA22-074A from the FBI and CISA, Russian state-sponsored actors gained access to a non-governmental organization's network as early as May 2021 [10]. They obtained the credentials through a brute-force attack on the password of an account that had a simple, predictable one [10].

That account had been unenrolled from the Duo two-factor authentication service after a long period of inactivity, but had not been disabled in Active Directory. Because Duo's default configuration allows a new device to be enrolled for a dormant account, the attackers registered their own, met the authentication requirements and got access to the network [10].

They then exploited the "PrintNightmare" flaw (CVE-2021-34527) to obtain administrator rights, then modified a file on the domain controller so that two-factor authentication calls were redirected to the local machine. Because Duo's default policy for Windows is to let logins through if the server is unreachable, two-factor authentication was in effect disabled [10]. They connected to the VPN, opened RDP sessions on the domain controllers, then reached the victim's cloud storage and email [10].

The advisory, published on 15 March 2022 and revised on 2 May 2022, recommends in particular enforcing two-factor authentication for everyone without exception while checking for default pass-through and re-enrollment risks, disabling inactive accounts uniformly in Active Directory and in the two-factor systems, setting delays and lockouts after repeated failures, and requiring strong, unique passwords [10].

## Sources

1. OWASP, "Blocking Brute Force Attacks". https://community.owasp.org/controls/Blocking_Brute_Force_Attacks (accessed 14 November 2026)
2. CNIL, "Mots de passe : recommandations pour maîtriser sa sécurité", 14 October 2022. https://www.cnil.fr/fr/mots-de-passe-recommandations-pour-maitriser-sa-securite (accessed 14 November 2026)
3. Proofpoint, "What Is a Brute-Force Attack?". https://www.proofpoint.com/us/threat-reference/brute-force-attack (accessed 14 November 2026)
4. NCSC (UK), "Password policy: updating your approach". https://www.ncsc.gov.uk/collection/passwords/updating-your-approach (accessed 14 November 2026)
5. OWASP, "Brute Force Attack". https://community.owasp.org/attacks/Brute_force_attack (accessed 14 November 2026)
6. NIST, "NIST Special Publication 800-63B". https://pages.nist.gov/800-63-4/sp800-63b.html (accessed 14 November 2026)
7. MITRE ATT&CK, "Brute Force, Technique T1110 - Enterprise | MITRE ATT&CK®". https://attack.mitre.org/techniques/T1110/ (accessed 14 November 2026)
8. MITRE ATT&CK, "Brute Force: Password Cracking, Sub-technique T1110.002 - Enterprise | MITRE ATT&CK®". https://attack.mitre.org/techniques/T1110/002/ (accessed 14 November 2026)
9. MITRE ATT&CK, "Brute Force: Password Guessing, Sub-technique T1110.001 - Enterprise | MITRE ATT&CK®". https://attack.mitre.org/techniques/T1110/001/ (accessed 14 November 2026)
10. CISA, "Russian State-Sponsored Cyber Actors Gain Network Access by Exploiting Default Multifactor Authentication Protocols and “PrintNightmare” Vulnerability", 15 March 2022. https://www.cisa.gov/news-events/cybersecurity-advisories/aa22-074a (accessed 14 November 2026)

---

Original source: Cyberpedia, the public cybersecurity and AI encyclopedia published by Brain. Canonical version: https://startbrain.io/cyberpedia/en/articles/dictionary-attack-mutation-rules-passwords/
You may reuse this text freely under the CC BY 4.0 license (https://creativecommons.org/licenses/by/4.0/), provided you credit "Brain, Cyberpedia" with a link to the canonical URL above.
