Can LLMs replace static analysis?
The answer to that question (at least today) is a resounding no - but our AI team at Semgrep has found that there is immense power in combining the two.
By adding AI post-processing to Semgrep scans, we’ve been able to eliminate a large number of false positives that static analysis tools alone will always flag. These are the false positives that require contextual understanding of code and frameworks to identify, which LLMs are really good at doing when paired with Semgrep’s SAST engine.
In this webinar, you’ll learn:
How Semgrep Assistant catches the false positives that static analysis alone could never, with a 95% human agree rate
How this reduction in noise equates to an average 20% reduction in findings to triage (imagine cutting your backlog by 20% overnight)
Why this noise filtering capability finally makes shift left a real possibility for AppSec teams and developers
Join us as we explore how Semgrep Assistant has evolved from providing insights in 2024 to becoming a more autonomous, agentic solution in 2025 and beyond!