Physical and Cyber Security

Physical and Cyber Security

AI applied to both physical security (access control, surveillance, anomaly detection) and cybersecurity (threat detection, penetration testing, incident response, compliance). Also covers the new attack surfaces that AI systems themselves introduce.

Key Topics

This domain is in its early stages. The security intersection with AI runs in both directions — AI as a tool for security, and AI systems as new attack surfaces:

• AI-powered threat detection — Using LLMs and agentic workflows for anomaly detection, log analysis, and automated incident response. Pages forthcoming.
• AI Agents — The agentic paradigm has direct security implications: autonomous systems that can interact with external tools present both defensive opportunities (automated pentesting, compliance checking) and offensive risks (agent hijacking, prompt injection attacks).
• AI system security — The unique vulnerabilities introduced by deploying LLMs in production: prompt injection, data poisoning, model extraction, and alignment failures. Pages forthcoming.

This domain will expand as we ingest security-specific sources. See the Index for the current state of all wiki pages.

Pages in This Domain