# Neurosecurity
**Entity class:** Security field
## Definition
**Neurosecurity** protects neural devices, neural data, decoding models, stimulation pathways, command authority, and the resulting state of the biological endpoint. In a read-only interface, the dominant problem is the confidentiality and integrity of observation. In a writable or [[wiki/Brain-Computer Interfaces|bidirectional brain-computer interface]], security also governs whether an instruction may cross from computation into neural function.
## Security boundary
The protected object is the complete [[wiki/Cyber-Biological System|cyber-biological system]]: device hardware, firmware, software, communications, identity, permissions, machine-learning models, [[wiki/Neural Command Channel|neural command channel]], stimulation parameters, physiological response, and recovery state. The decisive questions are:
1. Is the device trustworthy?
2. Is the neural command channel trustworthy?
3. Has the biological endpoint remained within, or been restored to, a safe state?
## Operational architecture
The mature loop is [[wiki/Continuous Vulnerability Intelligence|continuous vulnerability intelligence]] → [[wiki/Human-Machine State Estimation|continuous human-machine state estimation]] → [[wiki/Predictive Exploit Detection|predictive exploit detection]] → rapid containment → trusted remediation → [[wiki/Verified Neural Restoration|verified restoration]]. This extends vulnerability management through the implant and into the state produced at the nervous system.
## Relationships
- **attack class:** [[wiki/Brainjacking|Brainjacking]].
- **clinical control loop:** [[wiki/Closed-Loop Neuromodulation|Closed-Loop Neuromodulation]].
- **collection:** [[collections/Neurotech|Neurotech]].
- **counterterrorism bridge:** [[wiki/Human CVE|Human CVE]] and [[collections/Terrorism, Counterterrorism, and the Intelligence Environment|Terrorism, Counterterrorism, and the Intelligence Environment]].
## Sources / Provenance
- Xinyu Jiang et al., [“Cybersecurity in neural interfaces: Survey and future trends”](https://pubmed.ncbi.nlm.nih.gov/37883851/), *Computers in Biology and Medicine* 167 (2023), covering data-, permission-, and model-level attacks across forward decoding and backward neuromodulation or stimulation.
- Laurie Pycroft et al., [“Brainjacking: Implant Security Issues in Invasive Neuromodulation”](https://pubmed.ncbi.nlm.nih.gov/27184896/), *World Neurosurgery* 92 (2016).
**As of:** 2026-09-23