Written by: Mio Kanzaki(Security Analyst (AI))Reviewed by: 千葉圭介Updated
Fuzzing is a technique that sends large volumes of random input to find faults. The idea is to give a program more input than a person would ever think to type. OWASP describes it as a software testing technique for finding bugs, vulnerabilities, and unexpected behavior. The CISA NICCS glossary describes a fuzzing tool as one that automatically enters large amounts of random or malformed data to find vulnerabilities, crashes, and unexpected behavior.
How does fuzzing work?
Picture testing a vending machine not only by pressing buttons in the right order but by inserting a coin the wrong way round or mashing the buttons, to see how it breaks. Fuzzing is a machine doing this rough kind of trying in large numbers.
According to OWASP, unexpected or malformed input is fed to the program automatically, and you see whether it stops or behaves strangely. The problems found include a buffer overflow (a fault where writing goes past a set boundary) and faults that lead to a service outage.
As an advantage, OWASP names that the test design is simple and carries no preconceptions about how the system works. It also explains that faults a human eye tends to miss can be found.
On the other hand, OWASP says fuzzing tends to find simple faults, and in tests from outside without knowledge of the inside, it is hard to judge how serious a found problem is. It supplements, and does not replace, human review of code and debugging, so it is one tool among several in an assessment. The overall picture of assessment is organized on the web application assessment page.
Why does fuzzing matter?
A site with input forms or its own API that you built yourself can have the fault that the program stops on unexpected input. Even on a ready-made CMS, a feature added later can cause the same thing.
Fuzzing is a way of trying such input-related faults widely with the power of a machine. The range it finds is limited, however. When you commission an assessment, it helps to ask beforehand what will be checked and by what method.
To ask what an assessment of your site would cover, contact us.
Related terms
- Buffer overflow: a fault where writing goes past a set boundary.
- Vulnerability: a weakness in software or settings that can be abused in an attack.
Next step
The kinds of assessment and how they proceed are organized on the web application assessment page. The free URL check shows only what is visible from outside. For how a check would proceed, contact us.
Frequently asked questions
If we run fuzzing, will every fault be found?
Not necessarily. OWASP says fuzzing tends to find simple faults. Combine it with other testing methods and a person's checks.
Do we have to run fuzzing ourselves?
Not necessarily. If you have written few programs yourself, it can be enough to confirm, when you commission an assessment, which methods will be used to check what.
Frequently asked questions
If we run fuzzing, will every fault be found?
Not necessarily. OWASP says fuzzing tends to find simple faults. Combine it with other testing methods and a person's checks.
Do we have to run fuzzing ourselves?
Not necessarily. If you have written few programs yourself, it can be enough to confirm, when you commission an assessment, which methods will be used to check what.