AI governance infrastructure

The governance
layer for
AI agents.

Any agent in. Compliant, verified agent out.
Framework-agnostic governance for regulated industries.

No noise. Updates when there's something worth saying. ✓ You're on the list. We'll be in touch.
0 of 23
Australian financial institutions reviewed by ASIC had a mechanism for consumers to challenge AI decisions — ASIC REP 798, 2024
40%
of enterprise applications will embed AI agents by end of 2026, up from less than 5% today — Gartner
€35M
maximum fine under the EU AI Act for high-risk AI system violations, enforceable from August 2026
The problem

Agents are being deployed faster than
they can be governed.

Compliance teams are being asked whether their AI systems are governed. Right now, there is no standard way to answer that question — what an agent is permitted to do, what data it can access, whether it complies with regulatory obligations, or how its behaviour is audited over time.

01
No permission verification
Agents access data and invoke tools without a documented, auditable record of what they were authorised to do.
02
No regulatory mapping
Most agents are deployed without being assessed against the ASIC, APRA, Privacy Act, or EU AI Act obligations that apply to them.
03
No portable compliance artefact
When a board, auditor, or regulator asks how an agent is governed, there is nothing to show them. No document, no certificate, no audit trail.
04
No neutral governance layer
Microsoft governs Microsoft agents. OpenAI governs OpenAI agents. No enterprise runs a single-vendor stack. No platform governs them all.
How it works

Register any agent.
Receive a verified one back.

Clearpath operates as a pass-through compliance layer — framework-agnostic, policy-driven, and designed to produce the artefact your board and regulators need.

Step 01
Intake & assessment
Register your agent — system prompt, tool permissions, data access scope, framework. Clearpath runs 47 automated checks against the policy library: PII scope, injection resistance, permission bounds, bias, explainability, contestability.
Step 02
Gap report & remediation
Violations flagged with regulatory citations and severity scores. Self-serve remediation playbooks guide your team through fixes — no consultant required for 80% of cases.
Step 03
Clearpath Verified
A portable Agent Card documenting how the agent is governed — regulatory mapping, active guardrails, assessment scores, known limitations. Board-ready. Regulator-ready. Audit-ready.
The output

The Clearpath Verified
Agent Card.

The core artefact of every assessment. A structured, versioned compliance document that proves governance exists — designed to be presented to boards, auditors, and regulators.

If the Agent Card becomes the standard way regulated industries document AI agent compliance, Clearpath becomes the category.

Retirement Advice Generator
Example Financial Services Pty Ltd · AGT-2026-0041
Clearpath Verified
High risk LangChain GPT-4o ASIC licensed Certified 14 Mar 2026
Regulatory mapping
Policies applied ASIC RG 255 · Privacy Act · APRA CPS 230/234 · FAR
Accountable person Jane Smith, CCO
Data access Client records · Morningstar API
94
PII handling
88
Injection
76
Fairness
100
Audit trail
Active guardrails
Allocation caps✓ Enforced
RG 255 disclosure✓ Auto-generated
Email send gate✓ Human confirm required
Contestability path✓ Documented and tested
CP-2026-AFS-0041-V1 · SHA: a3f9c2b Clearpath Verified ✓
Who it's for

Built for the people
who answer to regulators.

Compliance officers
A clear answer when ASIC asks.
Produce a board-ready Agent Card that documents exactly how each AI agent is governed — without a manual audit process or an external consultant.
Risk managers
Governance before deployment, not after.
Assess agents before they go live. Catch violations before they become incidents. Continuous runtime monitoring flags anomalies before they reach customers.
Senior PMs & CTOs
Build fast. Deploy with confidence.
Framework-agnostic governance that works with your existing agent stack — LangChain, Copilot Studio, OpenAI, custom Python. No re-platforming required.
Regulatory coverage

Built for the frameworks
that govern you.

Clearpath's policy library maps to the specific obligations your agents must meet — Australian financial services first, with global frameworks in the pipeline.

Australian Privacy Principles APRA CPS 230 APRA CPS 234 ASIC REP 798 Financial Accountability Regime EU AI Act — coming ISO / IEC 42001 — coming NIST AI RMF — coming GDPR — coming

Your agents are
already deployed.
Are they governed?

Join the waitlist for early access. We're onboarding a small cohort of Australian financial services firms to run the first assessments.

✓ You're on the list. We'll be in touch shortly.
No pitch decks. No cold calls. Just a conversation when we're ready.