Every AI Policy Ever · open in the interactive register

Revoked / withdrawnProposed · Guidelines / framework

Introducing mandatory guardrails for AI in high-risk settings: proposals paper

Record AUS-04

Issued by
Department of Industry, Science and Resources
Jurisdiction
Australia
Level · region
National · Oceania
Published / adopted
2024-09-05
Language
English

Summary

Consultation paper proposing 10 mandatory guardrails for organisations developing or deploying AI in high-risk settings: accountability processes, risk management, data governance, testing and monitoring, human oversight, informing end-users, ways to challenge AI outcomes, supply-chain transparency, record-keeping and conformity assessment. It sought views on principles for defining high-risk AI and their application to general-purpose AI models, and on 3 regulatory options: adapting existing frameworks, framework legislation, or a new cross-economy AI Act.

Safety topics

risk-based classificationgeneral-purpose / frontier AIaccountability & liabilityevaluation & testinghuman oversighttransparency & labelling

Status notes

Published in September 2024; consultation ran from 5 September to 4 October 2024, and public responses were published. The paper followed the government's interim response to the Safe and Responsible AI in Australia discussion paper, and its guardrails were consistent with the Voluntary AI Safety Standard released for businesses in the meantime. The government later announced that it will not proceed at this time with the proposals to introduce mandatory guardrails; feedback on the paper informed the National AI Plan of 2 December 2025, which builds on existing, largely technology-neutral legal frameworks, with regulators keeping responsibility for AI harms in their domains, and establishes an AI Safety Institute.

Sources

Open the official source https://consult.industry.gov.au/ai-mandatory-guardrails

Local copy

Archived. Saved with a headless browser because the plain download looked incomplete. A source changed since the previous snapshot; both versions are kept. Last attempt 2026-10-17 21:35:45 UTC

Official source (third-party copy, not published) archive/AUS-04/2026-10-11/official.html
Retrieved 2026-10-11 13:59:15 UTC
SHA-256 9b83b3693f3c9242ff5c3a7f38bd81c293eb767177243084e2e26c53deb84f4f

Warning: The page is a JavaScript application shell; the saved copy holds only the site's navigation, not the document

Official source, rendered in a browser (third-party copy, not published) archive/AUS-04/2026-10-11/official-rendered.html
Retrieved 2026-10-11 14:10:57 UTC
SHA-256 dc0dd359d0e4eac1825e82b8db73cd146c77dd5d43e6a60178024a151cd4ba97
Official source, printed to PDF by a browser (third-party copy, not published) archive/AUS-04/2026-10-17/official-print-2.pdf
Retrieved 2026-10-17 21:37:22 UTC
SHA-256 0dfcd3b39a5244221ffd3b047a7118407f645fb23e5a56c40b789256a4638079

Verification

A
The instrument's own official text, or an official register, was stored and confirms the title, date and status. checked 2026-10-17 · official-url-fetched
Recheck of 17 Oct 2026: re-read the DISR consultation hub page as rendered in a browser (stored 11 Oct 2026) and as printed to PDF (stored again 17 Oct 2026, same text): title, department, consultation dates, the questions consulted on and the update that the Australian Government will not proceed at this time with mandatory guardrails. The plain download of the page is a JavaScript shell. Stored in this run: the proposals paper (September 2024), linked from the consultation page but hosted on a generic cloud storage host (storage.googleapis.com), so its quotes count as secondary; and DISR's National AI Plan page and PDF (2 December 2025). The consultation page gives no date for the decision not to proceed and the Plan does not mention the guardrails, so date_last_updated stays empty; the earlier notes' statement that it used the National AI Plan release date was removed.
Second check: Independent check by Claude Code (claude-sonnet-5-5), 2026-10-17: 7 field(s) supported.

History

  1. 2026-10-17notes: Consultation ran 5 Sep-4 Oct 2024. The government has since said it 'will not pr → Published in September 2024; consultation ran from 5 September to 4 October 2024 — rewritten from the stored consultation page, the paper and the National AI Plan: dates, the interim response it followed, the link to the Voluntary AI Safety Standard and the outcome; the sentence on date_last_updated contradicted the record (the field is empty)2026-10-17-recheck-europe-3 · Claude Code (claude-opus-5-5)
  2. 2026-10-17summary: Consultation paper proposing mandatory guardrails for developing and deploying A → Consultation paper proposing 10 mandatory guardrails for organisations developin — rewritten from the stored proposals paper: the 10 proposed guardrails, the principles for defining high-risk AI and their application to general-purpose AI, and the 3 regulatory options2026-10-17-recheck-europe-3 · Claude Code (claude-opus-5-5)

Related records

Australia National AI Plan Australia · Published (non-binding, current)

Australia VAISS (10 guardrails) Australia · Superseded

Citation

Department of Industry, Science and Resources. Introducing mandatory guardrails for AI in high-risk settings: proposals paper. 2024-09-05. https://consult.industry.gov.au/ai-mandatory-guardrails. Status: Revoked / withdrawn; verified 2026-10-17 (level A). Local copy archive/AUS-04/2026-10-11/official.html retrieved 2026-10-11 13:59:15 UTC, SHA-256 9b83b3693f3c9242ff5c3a7f38bd81c293eb767177243084e2e26c53deb84f4f. Every AI Policy Ever, version 1.18 (snapshot 2026-10-18), record AUS-04.

Improve this record

Know of an amendment, a status change or an error? Report news or an error on GitHub; the form opens with this record filled in.