"False negative -- that one is going to kill you": Understanding Industry Perspectives of Static Analysis based Security Testing
Not on the map. Why?
Abstract (arXiv)
The demand for automated security analysis techniques, such as static analysis based security testing (SAST) tools continues to increase. To develop SASTs that are effectively leveraged by developers for finding vulnerabilities, researchers and tool designers must understand how developers perceive, select, and use SASTs, what they expect from the tools, whether they know of the limitations of the tools, and how they address those limitations. This paper describes a qualitative study that explores the assumptions, expectations, beliefs, and challenges experienced by developers who use SASTs. We perform in-depth, semi-structured interviews with 20 practitioners who possess a diverse range of software development expertise, as well as a variety of unique security, product, and organizational backgrounds. We identify 17 key findings that shed light on developer perceptions and desires related to SASTs, and also expose gaps in the status quo - challenging long-held beliefs in SAST design priorities. Finally, we provide concrete future directions for researchers and practitioners rooted in an analysis of our findings.
Key facts. Cited by 8 papers in the map (data through 14 Oct 2026). First posted on arXiv 30 Jul 2023.
Builds on
No reference list yet.
Built on by
Citing papers per year: 2024: 1; 2025: 6; 2026: 1.
8 papers.
- 2025BaxBench: Can LLMs Generate Correct and Secure Backends?cited by 41
- 2025Large Language Models Are Unreliable for Cyber Threat Intelligencecited by 5
- 2025AutoBaxBuilder: Bootstrapping Code Security Benchmarkingcited by 5
- 2026Reducing False Positives in Static Bug Detection with LLMs: An Empirical Study in Industrycited by 5
- 2024Risks of ignoring uncertainty propagation in AI-augmented security pipelinescited by 1
- 2025Towards Effective Complementary Security Analysis using Large Language Modelscited by 1
- 2025MulVuln: Enhancing Pre-trained LMs with Shared and Language-Specific Knowledge for Multilingual Vulnerability Detectioncited by 0
- 2025Argus: A Multi-Agent Sensitive Information Leakage Detection Framework Based on Hierarchical Reference Relationshipscited by 0
- 2026Reducing False Positives in Static Bug Detection with LLMs: An Empirical Study in Industry26 Jan 2026
- 2025AutoBaxBuilder: Bootstrapping Code Security Benchmarking24 Dec 2025
- 2025Argus: A Multi-Agent Sensitive Information Leakage Detection Framework Based on Hierarchical Reference Relationships9 Dec 2025
- 2025MulVuln: Enhancing Pre-trained LMs with Shared and Language-Specific Knowledge for Multilingual Vulnerability Detection5 Oct 2025
- 2025Towards Effective Complementary Security Analysis using Large Language Models20 Jun 2025
- 2025Large Language Models Are Unreliable for Cyber Threat Intelligence29 Mar 2025
- 2025BaxBench: Can LLMs Generate Correct and Secure Backends?17 Feb 2025
- 2024Risks of ignoring uncertainty propagation in AI-augmented security pipelines14 Jul 2024