In short
Credential stuffing involves automatically testing, on a service, usernames and passwords stolen in data breaches on other sites; it exploits the reuse of passwords. It is a type of brute-force attack, along with password guessing and password spraying, which attempts a few very common passwords on a large number of accounts. Multi-factor authentication and the detection of automated logins are the main defenses against it.
The family of brute-force attacks
| Attack | Principle | What it exploits |
|---|---|---|
| Classic brute force | Trying all possible combinations, or a dictionary, on an account | Short or common passwords |
| Password spraying | Trying a few very common passwords on thousands of accounts | The absence of MFA and lockout thresholds calculated on a per-account basis |
| Credential stuffing | Reusing credentials obtained from data breaches on other services | Password Reuse |
| Offline cracking | Recovering passwords from their stolen hashes | Storage with a hash function that is too fast |
Password spraying is designed to bypass lockout thresholds: one or two attempts per account do not trigger a lockout.
Defenses That Work
- Prioritize multi-factor authentication for email, VPN, and administrative access.
- Reject passwords already found in known data breaches, rather than imposing arbitrary complexity rules.
- Limit the rate of login attempts per account, per IP address, and globally, and detect automated logins.
- Monitor for warning signs: spikes in failed login attempts, logins from unusual networks or countries.
- Store passwords using an appropriate hashing function (Argon2, bcrypt, scrypt) to make cracking them computationally expensive.
- Do not reveal in error messages whether a username exists, whether during login or password reset.
Frequently asked questions
How can I tell if my login credentials have been leaked?
Services like Have I Been Pwned indicate whether an email address appears in known public data breaches. If so, change the affected password wherever it was reused and enable multi-factor authentication.
Sources
Find out where an attacker would get in.
Our AI agents actually attack your scope, fully automated, and hand you the proof of every vulnerability.
