Detecting Sql Injection Vulnerabilities Using Artificial Bee Colony and Ant Colony Optimization

Due to the complexity and interconnection of many systems, nowadays is hardly to affirm that a system is secure. Thus, any tool, process or awareness that can improve security is very helpful. The Open Web Application Security listed “Injection” as the third most security risk for web applications in 2021. There are many automated tools to assist professionals, in order to identify this vulnerability. However, keeping these tools up to date has proven to be a challenge. Thus, there has been some interest in applying Artificial Intelligence (AI) in this field. In this paper, we propose an approach to detect SQL injection vulnerabilities in the source code, using swarm-based algorithms, Artificial Bee Colony (ABC) and Ant Colony Optimization (ACO). To test this approach empirically we used web applications purposefully vulnerable as Bricks, bWAPP, and Twitterlike. Simulation results verify the effectiveness of the ABC and ACO when compared with a static analysis tool.