AI fuzzing: A testing method to identify potential security vulnerabilities

All software comes with security vulnerabilities. To ensure the software is safe to use, a repetitive process of finding and patching these vulnerabilities is required. Let’s find out more about “AI fuzzing”, a testing method used to find security vulnerabilities. ‘Fuzzing’: A software security testing method

Fuzzing is one of many methods used to test software vulnerabilities. Random inputs are made into the software, and any errors or conflicts are monitored to find weak spots in security. Fuzzing is accepted among security experts as a rather effective testing method, as fuzzing is able to find more bugs than can found by analyzing source code. Sometimes while inputting variables for fuzzing, manual settings can be required. But inputting all possible variables is a very time-consuming, expensive, and practically impossible process. For this reason, fuzzing is called a “dumb science” by some. Fuzzing meets AI

However, traditional fuzzing combined with AI (artificial intelligence) is a completely different beast. AI fuzzing uses a technique similar to machine learning to identify security vulnerabilities in an application or system. Machine learning is an area of artificial intelligence, and refers to techniques used by a computer to analyze patterns and identify meaningful information in given data. Machine learning makes the generation of test cases for fuzzing much easier, and makes customized fuzzing methods possible. Vulnerabilities undetectable using conventional methods can now be found faster. AI fuzzing: Named among the world’s top 10 security threats in 2019

Nonetheless, while fuzzing techniques were married to AI for more effective security, Australian technology consulting firm “Rightsize Technology” has named AI fuzzing among the world’s 10 greatest security threats of 2019. Although AI fuzzing makes finding security vulnerabilities easier, this also means that it’s easier to carry out “Zero Day” attacks where security vulnerabilities are identified by an attacker and attacked before they can be patched. It’s often the case that more widespread use of a new technology also increases the chances of its abuse. AI technology promises to continue improving fuzzing techniques to find potential security vulnerabilities. We’ll need to keep a close eye on how fuzzing develops.