An AI governance platform that can actually stop something
An AI governance platform that decides whether an agent’s action is allowed, enforces the decision in the request path, and produces the record as a by-product. Inventory, ownership, policy and evidence in one control plane.
The problem
What this is actually about
Most platforms sold as AI governance sit beside the system they govern. They hold an inventory somebody maintains, a control mapping somebody assembled, and an evidence pack somebody exported before an audit. None of that is in the path of an action, so none of it can prevent one.
The test that separates the categories is blunt: can it deny an action in flight, and is its evidence produced by that decision or written down afterwards. A platform that answers no to both is a documentation product, which is a legitimate thing to buy and a different thing from a control.
What the platform does about it
Read from the product architecture rather than written here, so this page cannot promise something no module ships.
What AI agents exist here?
- Connectors across AI platforms, cloud, identity providers, source control, enterprise applications and infrastructure
- Shadow agent discovery from model API traffic, AI keys, MCP configuration, SDK dependencies and automation credentials
- An inventory record per agent: owner, purpose, model, framework, tools, MCP servers, data sources, permissions and secrets
- Ten lifecycle statuses from Discovered through Quarantined and Retired
Under which rule was this decided?
- Rules authored in natural language, compiled to structured policy, and verified by a human before activation
- Conditions over agent, owner, department, tool, action, resource, data class, geography, environment, time, risk, user, transaction value and tenant
- Simulation against historical traffic, so a rule states what it would have blocked before it blocks anything
- Versioned policy, with every decision recording the version that produced it
What is it doing right now, and should this action be allowed?
- A gateway between the agent and the tool, API or database it is calling
- A recorded decision per action, carrying agent, user, resource, context, risk, policy version and outcome
- Behavioural baselines, so a hundredfold jump in record reads reads as an anomaly rather than a busy afternoon
- Runaway protection that suspends an agent stuck in a loop before it exhausts cost or capacity
- Human approval for sensitive actions, and two-person approval for critical ones
What is the record, and can it be trusted?
- An append-only record of agent, user, session, resource, action, policy, decision, tool, model, result and risk
- Customer-configurable retention from 30 days to seven years
- Tenant isolation on every record, so no agent can read across the boundary
- Event-driven telemetry that keeps the authorization path fast and synchronous while analytics runs behind it
What it leaves behind
The artefacts these modules produce. Evidence generated by a decision, rather than assembled for an audit afterwards.
- A first-day discovery report naming agents nobody had registered
- An unregistered-agent alert with provider, data reach and risk
- A simulated impact figure before activation
- A policy version stamped on every decision
- An allow, deny or escalate decision on every governed action
- A suspension when behaviour leaves the baseline
- The evidence every other module cites
- A retention position the auditor can check
See it working
Every claim above has somewhere on this site you can go and check it.
- Watch a decision being made
An indirect prompt injection denied and a high-value refund held for a person, decision by decision.
- A worked estate
Twelve agents ranked by risk, every score broken into the factors that produced it.
- The record it leaves
Every event the estate produced, derived from its own decisions rather than asserted.
Common questions
- What makes this a platform rather than a policy document?
- Three things have to be true: an action passes through a decision point, a policy is evaluated there using current context, and the outcome is enforced rather than recommended. Without the first there is no control, only monitoring after the fact.
- Does it replace our AI governance documentation tooling?
- Usually not. Documentation platforms demonstrate that a programme exists and map controls to frameworks; a control plane decides and enforces at the moment of action. They are complements, and the mistake is treating them as alternatives.
- What does it need to see before it can enforce anything?
- An inventory, including the agents nobody registered. Every control downstream — ownership, least authority, approval, evidence — depends on knowing which agents exist and what authority they hold, which is why discovery runs before enforcement.
- Is enforcement all-or-nothing?
- No. A rule can run in monitor mode, which evaluates it and records what it would have done without changing the outcome, so a control can be measured against real traffic before it is switched on.