Arbitrator

    Policy enforcement and decision arbitration for autonomous AI agents. Arbitrator sits between an agent's intent and its execution, deciding — in real time — whether an action is allowed, denied, or escalated for human review.

    What Arbitrator Is

    As enterprises deploy autonomous and semi-autonomous AI agents into procurement, finance, customer support, DevOps, and internal tooling, a new question emerges that traditional IAM and DLP tools were never built to answer: is this specific agent action, right now, in this context, allowed?

    Arbitrator is Andeavour's decision-arbitration layer for agentic systems. It intercepts the actions an AI agent intends to take — API calls, database writes, fund transfers, code merges, ticket closures — and arbitrates them against a live enterprise policy model before execution is permitted. Unlike static access control, Arbitrator evaluates the full decision context: the requesting agent's identity and permissions, the action's blast radius, prior agent behavior, and any conflicting policies triggered by the request.

    Arbitrator is deployed alongside AgentGuard and the broader Agentic SOC to give security and compliance teams a single arbitration point across every agent action, regardless of which model, framework, or vendor built the agent.

    How Arbitration Works

    1
    Agent submits an intended action to Arbitrator's decision API before execution
    2
    Arbitrator evaluates the action against policy, agent identity, and behavioral risk score
    3
    Conflicting policies are resolved by precedence rules and severity weighting
    4
    A verdict — allow, deny, throttle, or escalate — is returned in milliseconds
    5
    Every decision is logged with full context for audit and forensic replay

    The Policy Model

    Arbitrator's policy engine is built around three layers that most agent frameworks conflate into a single, brittle permission check. Separating them is what lets Arbitrator make defensible, auditable decisions instead of binary allow/deny calls.

    Capability Policy

    Defines what an agent or AI Principal is structurally permitted to do — scoped by identity, role, and the systems it has been granted access to.

    Contextual Policy

    Evaluates the specific request — dollar amount, data sensitivity, target system, time of day, requester chain — against dynamic risk thresholds.

    Precedence Rules

    When capability and contextual policies conflict — a common scenario in multi-agent workflows — precedence rules deterministically resolve the outcome instead of failing open.

    Deployment and Integration

    Inline Decision API

    Agent orchestration frameworks (LangGraph, CrewAI, AutoGen, custom orchestrators) call Arbitrator's decision API before executing tool calls, ensuring nothing runs without an approved verdict.

    Policy-as-Code

    Security and compliance teams author policies as version-controlled code, reviewed through the same pipelines used for infrastructure-as-code changes.

    Zero Standing Trust

    Every action is arbitrated independently — no agent inherits blanket approval from a prior decision, closing the gap that lets a compromised agent chain privileged actions together.

    Works With Your Platform

    Arbitrator plugs into the same telemetry and identity graph used across the Andeavour platform, so arbitration decisions inform detection and vice versa.

    Use Cases

    Finance Agents

    Arbitrate spend-approval agents against dollar thresholds, vendor allowlists, and dual-control requirements before a payment API is ever called.

    DevOps and Code Agents

    Block or escalate autonomous code-merge and infrastructure-change agents that touch production, secrets, or IAM configuration.

    Customer-Facing Agents

    Prevent support and sales agents from issuing refunds, discounts, or account changes outside pre-approved bounds — with human escalation for edge cases.

    Frequently Asked Questions

    How is Arbitrator different from an IAM permission check?

    IAM answers "is this identity allowed to call this API at all." Arbitrator answers "should this specific action, by this agent, in this context, execute right now" — factoring in behavior, blast radius, and policy conflicts that IAM systems don't model.

    Does Arbitrator add latency to agent workflows?

    Arbitration decisions typically return in single-digit milliseconds. High-risk or ambiguous actions can be routed to asynchronous human review without blocking lower-risk agent activity.

    What happens when two policies conflict?

    Precedence rules — configured by your security and compliance teams — resolve conflicts deterministically. Arbitrator never fails open by default; ambiguous cases are escalated rather than silently allowed.

    Can Arbitrator run alongside AgentGuard and AgentHunter?

    Yes. Arbitrator enforces policy on internal agent actions, while AgentGuard secures the AI agents and Principals themselves, and AgentHunter detects agentic attacks originating outside your perimeter. Together they form the enforcement, identity, and detection layers of the Andeavour platform.

    Where can I learn more about the underlying architecture?

    See how the Andeavour platform works and what agentic AI security means for a deeper technical grounding, or browse our research on real-world agentic attack patterns Arbitrator is designed to stop.