Governs sits between your organisation and every AI model it uses. Shadow AI is logged and classified. Australian PII never leaves your network. AIAF compliance assessments populate themselves.
Deploys in 30 minutes. Prompts never leave your network.
The average number of staff at an Australian government agency calling ChatGPT, Claude, or Gemini directly from work devices right now — some of them pasting Tax File Numbers, Medicare numbers, and classified documents into foreign AI systems. Nobody knows. Nobody's stopping it. And nobody can prove it didn't happen.
Everything works from the moment you point your first application at the gateway.
A DNS sinkhole intercepts direct calls to ChatGPT, Claude, Gemini, and other public AI services. Every attempt is logged with full context: who, what service, timestamp, and PII classification. Shadow AI becomes a managed risk with an audit trail — not an invisible liability.
Instead of blocking and losing the user, Governs surfaces each incident to your CISO dashboard. The conversation that follows — "you're using AI, let's bring it under governance" — is your upsell moment.
Every prompt passing through the Governs gateway is scanned for Australian PII before it reaches the upstream model. Tax File Numbers, Medicare numbers, BSB combinations, ABNs, passport numbers, and state-specific driver's licences — all built to the patterns that the Privacy Act 1988 actually protects.
Three enforcement modes: flag and alert, redact before forwarding, or block and return an error. Per-application, per-team, or organisation-wide policy.
Router observes every AI system and auto-populates up to 70% of your Australian AI Assessment Framework assessment from trace data. Nine dimensions filled automatically. Your compliance officer reviews the rest and signs off. The report writes itself.
Real-time spend by team and model. Every request scored for complexity — Governs shows exactly which calls are running on expensive models unnecessarily, what you'd save by routing them down, and creates the routing rule in one click.
Every AI agent's reasoning chain visualised in real time. Tool calls, LLM invocations, and decision points as they happen. Built on OpenTelemetry spans. Your governance team sees what an agent actually did — not just that it ran.
Not a US product with GDPR templates bolted on. Every feature is designed around Australian obligations from the ground up.
One Docker container. One DNS configuration. One line change per application.
Run the Governs container inside your network. It starts an OpenAI-compatible gateway and sets up the DNS intercept for shadow AI detection.
Point your existing AI applications at the Governs gateway. The OpenAI SDK, LangChain, and LlamaIndex all work without any other modifications.
Every AI call — authorised or shadow — is now visible, PII-scanned, cost-tracked, and AIAF-assessed. The dashboard is live within minutes of first traffic.
This is exactly what we run in every product demonstration — against your own AI traffic, not a scripted environment.
Point an existing OpenAI application at Governs. Nothing breaks. A trace appears in the dashboard immediately. The audience sees their own application, governed, in under a minute.
Live cost dashboard. AI spend broken down by team and model. Most organisations have never seen this number before and it is almost always larger than expected.
The ROI panel. Typically $2,000–$4,000 per month in identified routing savings for a mid-size agency. One click creates the routing rule. The audience creates it themselves.
A shadow IT event triggered live. PII flagged. Employee identity, team, timestamp, and prompt preview — all logged in real time on the CISO dashboard.
Auto-populated from observed behaviour. The compliance officer opens a near-complete AIAF assessment based on five minutes of live traffic. They review the remaining fields. The report writes itself.
Pilot programs available for government agencies. Contact us for volume pricing.
Per AI system · per month
Contact us for pricing
Everything in Govern, plus
Three-layer shadow AI coverage
For agencies requiring dedicated infrastructure
Custom pricing
We'll run the demonstration against your own AI tools in five minutes. No slide deck. No scripted environment. Your AI, governed — live in the room.