Obsignor · Australia’s AI-decision record
Know every AI tool you use, and show you’re using it responsibly.
Obsignor keeps your AI register, checks the risk of each use with fixed rules (no AI guessing), and writes what Australian privacy law now asks for, including the automated-decision wording your privacy policy needs from 10 December 2026. Every evidence pack is sealed, so no one can quietly change it, and ready for your board, the OAIC, your insurer, or a client who asks.
Built for organisations of about 20 to 2,000 people without a risk department. See pricing or book a 25-min research call.
Built to Australian rules: the Privacy Act (including APP 1.7–1.9), the statutory privacy tort, the Guidance for AI Adoption and the Voluntary AI Safety Standard, with packs for health, professional services and financial services.
Someone has to answer for your AI. Usually it’s you, on top of your day job.
No list
Staff are already using AI tools nobody approved. You find out afterwards.
No time
Doing it properly means a policy, a register, a risk check per tool and privacy wording, all by hand in Word.
No risk department
Enterprise governance tools assume a team you don't have, and treat Australia as an afterthought.
Decide which AI to build. Govern the ones you do.
Two tools on one spine. Start in Foundry — your first evaluation is free, no account needed — to decide. A use case you pursue flows straight into the Register, already scoped, to govern.
Decide
Should we build this AI?
Govern
Is it safe, evidenced, board-ready?
organisation context · deterministic triage engine · audit & export
Foundry — Decide
FreeShould you build this AI at all? Describe it in plain English and get a decision-grade read in a minute — a Go / Refine / Park / Stop verdict, an order-of-magnitude cost to own, model guidance, and a preview of its governance risk.
- Value, feasibility & viability scored
- Cost to own & model guidance
- Deterministic — every score shows its reasoning
Register — Govern
Keep the record of every AI tool you use: a risk check with fixed rules, the assessments that level of risk needs (drafted for a person to review and own), and a sealed export you can show anyone who asks, with a named person accountable for every approval.
- Risk check — Low / Medium / High, reasons shown
- AI drafts; a named person approves
- Sealed export for your board, auditor or clients
Both share your organisation’s details, the same risk rules and the same record, so Foundry’s risk preview and the Register’s full assessment always tell one story.
Three reasons to get your AI on record now.
None of them needs a risk department to apply to you, and none of them goes away after one date.
Privacy Act · from 10 December 2026
Say how you use automated decisions
If a computer program makes, or substantially helps make, decisions that significantly affect people, your privacy policy must say so. Applies to most organisations with turnover over $3 million, and to health providers of any size.
Statutory privacy tort · in force now
People can sue for serious invasions of privacy
Watching or recording people, biometrics and misusing their information are where AI gets organisations into trouble. The tort applies whatever your size.
The questions you'll be asked
What AI do you use, and how do you control it?
Clients’ supplier questionnaires, insurers and boards increasingly ask. Have the answer ready, with the record behind it.
APRA-regulated or listed? See Obsignor for financial services for CPS 230/234 and APRA’s 2026 letter on AI.
One record. Everyone who needs it.
In most organisations there’s no AI-governance team, just the person who answers for it, the people who do the work, and the people who check.
Answers for it · Owner · CEO · COO · practice manager · general counsel
Show what AI you use and why it's safe, in one place.
One register of every AI tool, its risk, who approved it and what the privacy policy says about it. Ready when someone asks.
Does the work · Operations · IT · privacy · compliance
Stop governing AI in a folder of Word docs.
Import the register you've started, let Obsignor check the risk and draft the paperwork for you to edit, and export it in an afternoon.
Checks it · Board or committee · auditors · insurers · clients
A sealed record they can read.
Readers never cost a seat. Give an auditor time-limited, read-only access, or hand over the PDF.
Fixed rules where it must be explainable. AI where it saves you time. A person where someone must be accountable.
Built for Australia
The risk rules, controls and assessments are built to Australian law, not a global product with Australia added later.
Sealed when it's made
Each use case's evidence pack is sealed with a SHA-256 hash, so you can prove it wasn’t changed afterwards.
Risk you can explain
Risk levels come from fixed rules, with the reasons shown. The same answers always give the same result.
We govern our own AI
The platform is itself a high-tier AI use case under its own triage logic — and we govern it with the exact framework we sell.
Own the runtime? We’re the evidence layer next to it.
As AI risk lands on the CISO’s desk — CPS 234, agent oversight — Obsignor is the governance evidence layer alongside your security stack, not another runtime tool. Tamper-evident records, deterministic triage, and mappings to the standards your board asks about.
Early access
Join the waitlist.
We’re opening Obsignor to Australian organisations in small groups ahead of 10 December. Tell us about yours and we’ll be in touch as places open. No obligation, and you’ll get the readiness kit straight away.
- First access when your cohort opens
- Founder-led onboarding — help shape the roadmap
- The APP 1.7 readiness kit, immediately
Start with a use case.
Evaluate one in Foundry in a few minutes — the first is free, no account needed. See a worked example first if you like, or join the early-access list to govern it in the Register.
Also free: the APP 1.7 Readiness Kit.
Get the newsletter
Practical notes on deciding which AI to build and governing it well, for Australian teams. About once a month.
ob·sig·nor /ob-SIG-nor/ noun — Australian software that keeps the sealed record of an organisation’s AI decisions: what was decided, on what evidence, by whom, and when. From Latin obsignare, “to seal”.
Readiness aid, not legal advice — consult qualified Australian legal and regulatory advisers for compliance decisions.