# Human-Machine State Estimation **Entity class:** Predictive security method ## Definition **Human-machine state estimation** fuses observations from a person and connected computational system to infer their joint, partially hidden operating state. In a neural interface this may include device configuration, software and model versions, communication integrity, identity and permission state, stimulation history, neural telemetry, physiological response, and uncertainty. ## Function The estimator does not merely ask whether a device passed a periodic security check. It continuously updates what is vulnerable, what changed, which attack paths became possible, which observations increase or decrease concern, and which intervention is most likely to restore an authorized safe state. ## Relationships - **input:** [[wiki/Continuous Vulnerability Intelligence|Continuous Vulnerability Intelligence]]. - **system:** [[wiki/Cyber-Biological System|Cyber-Biological System]]. - **output:** [[wiki/Predictive Exploit Detection|Predictive Exploit Detection]]. - **recovery reference:** [[wiki/Trusted Neural Baseline|Trusted Neural Baseline]]. - **mathematical relation:** [[wiki/Latent State Estimation|Latent State Estimation]]. ## Sources / Provenance - Xinyu Jiang et al., [“Cybersecurity in neural interfaces”](https://pubmed.ncbi.nlm.nih.gov/37883851/), 2023. - National Institute of Standards and Technology, [National Vulnerability Database](https://www.nist.gov/itl/nvd), machine-readable vulnerability enrichment and synchronization updates, accessed 2026-09-23. **As of:** 2026-09-23