Checkmarx’s new SAST layer adds AI‑driven triage, not just a bigger LLM
Checkmarx’s latest SAST release bundles a rule‑based scanner, a fine‑tuned security LLM, and a Findings Analysis Engine that filters true from false positives before developers see them. The trio claimed a 0.499 F1 score—over twice the category average—and uncovered dozens of bugs missed by leading frontier models. It signals a move from pure AI scanning to AI‑enhanced triage.