LATIDIA · Ciberseguridad

Dónde se silencian los arneses de descubrimiento de vulnerabilidades de IA: falsos negativos estructurales y vulnerabilidades ocultas

arXiv: 2610.29379v1Announce Type: new Resumen: Los arneses de descubrimiento de vulnerabilidades basados en LLM pueden pasar por alto vulnerabilidades reales y accesibles debido a la forma en que se construye el arnés alrededor del modelo. Pasos que seleccionan el código, lo filtran

La noticia

arXiv:2610.29379v1 Announce Type: new Abstract: LLM-based vulnerability discovery harnesses can miss real, reachable vulnerabilities because of how the harness around the model is built. Steps that select code, filter it, and store and merge findings can drop a vulnerability before or after the model runs, independently of the model's capability, and the report usually gives no sign that this happened. These drops come from engineering choices, and errors. They leave vulnerabilities undetected, and an attacker who knows the harness's rules can also use them to hide a vulnerability on purpose. Understanding these drops shows how to build harnesses that fix or disclose them and are harder to evade. Starting from