AI agent security for regulated and defense teams.
Keep the console in your own environment, keep content on the device, and decide what happens when MoorAI can’t decide.
- Management Servera Docker container in your environment
- What reaches ita category, a risk level, a keyed hash
- When MoorAI can’t decidethe administrator chose, per surface
- The agentopen source (MIT), so you can audit it
Your console, in your environment.
Run the MoorAI Management Server as a Docker container in your own environment, available on request. Self-hosted it is free, with no user limit, and source-available under the Elastic License 2.0, so the console that holds your fleet’s records stays inside your boundary.
Deployment options →Presets for regulated work.
The Regulated preset asks before money, identity and destructive changes, blocks bulk export and fails closed. The Industrial / OT preset blocks register, coil, setpoint, controller and safety-system writes by tool name and fails closed. It governs the agent’s calls on its own host, never inside a PLC, a safety system or the historian.
AI as critical infrastructure →Regulated
Industrial / OT
The administrator decides the fail mode.
When a check throws, runs out of time or can’t read its input, the administrator decides whether the call goes ahead or stops, per surface, in the signed policy or the root-owned machine config, never in a file the agent can write. Each fail-closed refusal raises a content-free alert.
How it works →Fail open
Fail closed
Evidence without the content.
Records carry a category, a risk level and a keyed one-way hash, never the prompt, and the console exports a signed evidence pack. MoorAI’s controls are mapped to the frameworks your auditors use. A mapping is evidence, not a certification.
Compliance crosswalk →Keep AI agents inside your rules.
Self-host the Management Server free, with no user limit, and enroll your devices.