SECURE CODE REVIEW WITH AI: ENHANCING STATIC AND DYNAMIC ANALYSIS

Over the years, as the software systems grew in complexity and connectivity, the risk of getting a vulnerability embedded in a piece of code has also significantly grown.Secure code review-a contemporary practice in software development-becomes the first and last line in defense of the cyberworld.Traditional code reviewing methods, which are attacked by way of static and dynamic analyses, have been useful but limited due to scaling issues, the false positives they cannot avoid, and the inability to adapt to changing threat landscapes.The review herein aims to elucidate the transformation of secure code review with AI-powered enhancement of both static and dynamic analysis techniques.AIpowered models, chiefly those of ML and NLP genres, have been contributing toward discovering vulnerabilities accurately, minimizing false-positive cases, and interpreting codes into further insights about behaviors.Inasmuch as the melding of AI with static tools such as CodeQL, SonarQube, and Semgrep and with dynamic tools such as AI-guided fuzzers turns the tide from reactive to proactive security, this paper focuses on the scientometric study of AI-assisted code review methods, tools, and research and attempts to answer issues of volume scarcity and integration.