Explore our new Cybersecurity Terms Glossary, explained by our experts.
Cybersecurity Terms Glossary
A brute force attack tries passwords, PINs, or keys one after another until one works. Learn how online and offline guessing differ and which controls stop them.

What Is a Brute Force Attack?
A brute force attack is an attempt to discover a password, PIN, or cryptographic key by systematically trying candidates until one works. It needs no flaw in the software—only time, computing power, and a target that accepts or exposes enough guesses. That simplicity is exactly why it remains one of the most common routes to account compromise.
The distinction that matters most to defenders is where the guessing happens.
In an online attack, every guess is sent to a live service: a Microsoft 365 sign-in page, an exposed Remote Desktop (RDP) port, an SSH daemon, a VPN portal, or an API login endpoint. The attacker is limited by network latency and by whatever throttling the service applies, and every failed attempt can leave a log entry. Tools such as Hydra and Medusa automate this against dozens of protocols.
In an offline attack, the attacker already holds password hashes—dumped from a breached database, extracted from a domain controller’s NTDS.dit file, or captured as a Wi-Fi handshake—and cracks them on their own hardware with Hashcat or John the Ripper. No lockout policy applies, nothing is logged on your side, and a single modern GPU can test billions of candidates per second against fast, unsalted hashes such as MD5 or NTLM. Here your only defence is how the passwords were stored and how long they are.
MITRE ATT&CK tracks this family as T1110 Brute Force, with four sub-techniques that map neatly to real-world behaviour:
Pure exhaustive search—every combination of every character—is rarely worth it beyond short PINs. Dictionary and hybrid attacks deliver most of the results, because people choose predictable passwords.
Each extra character multiplies the search space. An eight-character password drawn from all 95 printable ASCII characters gives roughly 6.6 quadrillion combinations, which sounds a lot until you remember that cracking rigs work through fast hashes at billions per second. Add four more characters and the figure grows by a factor of more than 80 million. That is why the NCSC recommends passphrases built from three random words, and why NIST SP 800-63B favours length, screening against known-breached passwords, and no forced periodic changes over arbitrary composition rules.
Online brute forcing is noisy if you are looking for it. Useful signals include:
Feeding these events into a SIEM with correlation rules turns them into alerts rather than log noise. Because attackers now distribute guesses across residential proxy networks, base your rules on the targeted account as well as the source address.
Our two-factor authentication solutions help organisations roll out MFA across VPN, email, and remote access without disrupting users.
It depends on where the guessing happens and how the password is stored. Online guessing against a throttled service may manage only a handful of attempts per minute, so a long, unique password is effectively safe. Offline cracking of fast hashes such as NTLM can test billions of guesses per second, so short or common passwords fall in seconds. Long passphrases stored with Argon2id or bcrypt can take impractically long to crack.
Lockout slows guessing against a single account, but it does not stop password spraying, where attackers try one password across many accounts and stay below the threshold. It also does nothing against offline cracking. Aggressive lockout can even be abused to lock legitimate users out. Combine sensible lockout or throttling with MFA, breached-password screening, and monitoring for failed logons spread across many accounts.
Testing systems you own or are explicitly authorised to test is lawful, and password audits are a normal part of security assessments. Attempting to access anyone else’s systems without permission is an offence under the Computer Misuse Act 1990. Even internally, get written approval, agree the scope, and warn the service desk, because testing can trigger lockouts and alerts that look like a real attack.
Start with authentication logs: Windows Security events 4625, 4624, and 4740 on domain controllers, Entra ID or Okta sign-in logs, VPN and firewall authentication logs, and web application login records. Alert on many failures followed by a success, many accounts failing from the same source, and any sign-ins over legacy protocols. Centralising these in a SIEM makes cross-system patterns visible.
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.