LATIDIA · Ciberseguridad

Enrutamiento de adaptadores con clave tokenizada: un mecanismo de control de acceso seguro contra la fuga de datos privados en LLM.

arXiv:2610.00309v1 Anuncio Tipo: nuevo Resumen: Los modelos de lenguaje grandes (LLM) se implementan cada vez más en dominios críticos para la privacidad (por ejemplo, atención médica, finanzas y gobierno), pero su propensión a memorizar y divulgar

La noticia

arXiv:2610.00309v1 Announce Type: new Abstract: Large language models (LLMs) are increasingly deployed in privacy-critical domains (e.g., healthcare, finance, and government), but their propensity to memorize and disclose personally identifiable information (PII) poses serious security and compliance risks. Existing defenses typically force a trade-off between model utility, privacy protection, and access to fine-tuned private knowledge. We propose LoRA-Oriented Control via Keyed Entry Tokens (Locket), a practical framework that embeds fine-grained, policy-driven access control directly into LLM generation. Locket trains a set of lightweight LoRA (Low-Rank Adaptation) adapters, each encoding a distinct access policy (e.g., full reveal, partial redaction via PII masking, or reveal under a specified differential privacy level). A compact