LATIDIA · Ciberseguridad

No juzgue a un agente por sus mensajes: fugas del Estado privado bajo monitoreo pasivo de la comunicación del agente

arXiv:2610.25032v1 Tipo de anuncio: nuevo Resumen: Un agente de IA puede reproducir la distribución de sus mensajes y de los resultados de la tarea del receptor mientras revela información privada que la tarea nunca requirió. Preguntamos qué

La noticia

arXiv:2610.25032v1 Announce Type: new Abstract: An AI agent can reproduce the distribution of its messages and of the receiver's task outcomes while disclosing private information that the task never required. We ask what a passive audit can certify when the auditor knows the authorised sender and the exact observable law but has no trusted pairing of messages with the private states behind them. For a single round, compatible mutual-information leakage forms a sharp interval, and fixing the entire task-relevant message law, with every declared task utility, can still leave a full private bit undetermined; total leakage is determined if and only if, in every task state, the remaining private state