The launch event · registration open
Webinar 5 · The Boundary, Drawn — launching the framework
Thu 24 Sep 2026 · 1:00–1:30 pm AEST · online, free, 0.5 CPD
Asset Management Council · Asset Zero
The Human–AI Boundary Framework for Asset Management
Where does AI end and human judgement begin? An Australian and New Zealand community answered — 2,796 recorded responses across four webinars — and a volunteer working group turned the answer into community-developed guidance, ratified unanimously and frozen as Version 1.0 on 31 August 2026: the ladder, seven principles, a decision-rights matrix from practitioners’ own votes, the boundary test, envelopes and the stop, records that make oversight verifiable. It stays living under AMC stewardship — v1.x changes are proposed in writing, answered in writing, closed on rough consensus.
How to cite
Asset Management Council (2026). The Human–AI Boundary Framework for Asset Management, v1.0. Community-developed guidance, frozen 31 August 2026. Available at projects.amcouncil.com.au/human-ai-boundary.
Cite a section by its identifier (e.g. “§B3”) — identifiers are stable across versions; page numbers are not.
Words from the Project Chair
Thank you to every volunteer who took part in the webinars and joined the working group. Version 1.0 is a minimum viable framework, and that is deliberate: AI moves in weeks, so we will iterate fast. Prediction, trend analysis and image recognition are yesterday’s AI — the real power arrives when AI drafts, reasons and acts inside human-set boundaries. This framework was built exactly that way: AI drafted, the community judged, a named human signed. Join the next iteration.
Two minutes, for AMCouncil members: how you’d take part, and what v1.x should focus on.
Built from what the community said
✓ Completed phase: Jun–Jul 2026
Webinar 1 · 25 Jun 2026
Knowing the Asset
663
responses
Webinar 2 · 1 Jul 2026
Condition, Performance and Risk
524
responses
Webinar 3 · 15 Jul 2026
Planning, Prioritisation and Investment
625 +28
responses
Webinar 4 · 30 Jul 2026
Work, Operations and Intervention
956
responses
Four webinars, then three volunteer review cycles in August — every comment answered in writing, dissent recorded and shipped, 162 final verdicts with zero objections. Walk the whole story →
The framework is living — what happens next
🧭
What’s next
The working group’s ranked recommendations — adopt the framework came first — the v1.x worklist, and where to have your say.
❓
The open questions
The honestly-held unknowns carried into v1.x, each with what would settle it.
💬
Propose a change
Every section carries a discussion door — signed-in AMC members propose v1.x changes; every input gets a written answer.
To our knowledge, the first project to pre-register and publish per-function human-AI boundary levels with enforcement telemetry — the project ran under the boundaries it wrote. How it works →