LATIDIA · Ciberseguridad

Conocido por sus acciones: toma de huellas dactilares de los agentes del navegador LLM a través de los rastros de la interfaz de usuario

arXiv: 2605.14786v2Tipo de anuncio: reemplazar Resumen: A medida que los agentes basados en LLM navegan cada vez más por la web en nombre de los usuarios, surge una pregunta natural: ¿pueden los sitios web identificar pasivamente qué modelo subyacente impulsa a un agente? Doi

La noticia

arXiv:2605.14786v2 Announce Type: replace Abstract: As LLM-based agents increasingly browse the web on users' behalf, a natural question arises: can websites passively identify which underlying model powers an agent? Doing so would represent a significant security risk, enabling targeted attacks tailored to known model vulnerabilities. Across 14 frontier LLMs and four web environments spanning information retrieval and shopping tasks, we show that an agent's actions and interaction timings, captured via a passive JavaScript tracker, are sufficient to identify the underlying model with up to 96\% F1 within a fixed harness. Although fingerprints are harness-specific, with cross-harness transfer falling to near chance, pooling traces across harnesses recovers model identification at 0.74