Security | Threat Detection | Cyberattacks | DevSecOps | Compliance

That is not a failure of fuzzing. It is a failure of interpretation. In a recent AFL++ fuzzing campaign targeting libarchive, we ran approximately 8.5 billion executions across all fuzzing phases, generated over a thousand crash files, and ultimately reduced them to two unique crash sites through structured crash triage and deduplication. This blog is a practical, engineering-first guide to that process: If your fuzzing pipeline stops at crash counts, you are not measuring security.

357 crash reports. 2 actual bugs. That is not a typo. That is the reality of modern application security testing. In a recent fuzzing campaign, over a thousand crash files were generated across billions of executions. After crash deduplication and triage, that number collapsed to just two unique issues. Not hundreds of vulnerabilities. Not dozens of risks. Two. And yet, most security teams would have celebrated the initial numbers.

Most mobile security workflows end in a familiar way. A scan runs, a report is generated, and the output looks reassuring. There are no critical issues, maybe a few medium findings, nothing that blocks a release. The process completes, the team moves forward, and the app ships. At that moment, the assumption is clear. The app has been tested. The risk is understood. But there is a question that rarely gets asked, and it changes the entire conversation.

“AI-powered DAST” is everywhere. It signals progress, but assumes something fundamental was missing. It wasn’t. DAST struggled not from lack of intelligence, but from lack of depth. Most tools never reached inside authenticated, stateful, multi-step journeys where real logic, sensitive data, and critical vulnerabilities exist. That’s the part Appknox solved years ago. AI here is not a reset. It is an accelerator, applied to a system already operating where risk actually lives.