Palo Alto Koi Agentic Endpoint Security

Endpoints are no longer made up of just an operating system and executables. The applications, open-source packages, browser extensions, local LLM models, MCP (Model Context Protocol) servers and skills that employees and AI agents keep adding form a new attack surface. Traditional EDR tools struggle to monitor threats that originate from this non-binary software; Koi Agentic Endpoint Security was built to close that blind spot.

The solution rests on four core capabilities. Total Visibility builds a real-time inventory of the applications, drivers, code packages, AI models, extensions, containers and skills on every endpoint, showing what data each component can reach, which external services it talks to and what actions it can take. Actionable Risks produces context-aware priorities with an LLM-driven risk engine that correlates reputation, exploitability, code intent and privilege boundaries. Proactive Guardrails and Control defines which packages, extensions and AI tools may be used, and preserves productivity by letting users request approval with a business justification when something is blocked. Real-Time Prevention places synchronous security hooks in the agent's execution path to stop indirect prompt injection, unauthorised tool calls and automated data exfiltration before they happen.

Palo Alto Networks completed its acquisition of Koi in April 2026. Koi Agentic Endpoint Security is offered both as a standalone solution and as a module integrated into Cortex XDR and Prisma AIRS: Cortex XDR gains a new module to find and remediate risk in the AI software ecosystem, while Prisma AIRS extends to agentic AI on the endpoint to provide a single control plane across the enterprise.