Submission 406
Mr Sunyal Maroo
Submitter group: Individuals.
Position labels are not shown for individual submitters.
An AI model identified the submitter group and chose the extracts. Counts do not measure importance or merit.
Extracts from the submission, in the submitter’s own words.
Extracts
Quotes are the submitter’s own words. Text taken from documents, including scanned pages read by text-recognition software, can differ from the PDF. Check the original. Listed in the same order as on the submission’s page in the interactive guide.
Australia's approach to AI governance relies on voluntary guidance at precisely the moment when those building frontier AI systems describe the risks as existential and call for independent oversight.
Australia's consumer protection, privacy and critical infrastructure laws are technology-neutral and will apply to harm once it occurs. But technology-neutral liability is a remedy after damage. It is not a preventive control.
The comparison is instructive because the UK is converting principles into binding obligations and Australia is not.
The gap is that these are internal policy instruments covering non-corporate Commonwealth entities, with defence and national intelligence carve-outs, rather than legislation creating rights individuals can rely on.
The National Cabinet outcome of 26 August 2026 endorsed mandatory national standards for large data centres covering energy, water, land use and skills. It does not cover data residency. This is a significant omission.
Extracts from recommendations
Extracts chosen by an AI model, in the submitter’s own words. Listed in the same order as on its page in the interactive guide.
The Australian Standards for AI framework should extend beyond data centre infrastructure to mandatory, enforceable obligations for developers and deployers of advanced AI systems, including pre-deployment safety assessment and independent audit for high-risk use cases.
The Australian AI Safety Institute should be obliged to report annually to Parliament on capabilities and risks assessed, including thresholds approached or exceeded.
Legislate the transparency obligations already operating in the Commonwealth Public Service as a public AI register, requiring publication of purpose, decision type, model or vendor, and impact assessment outcome for any AI system used by a Commonwealth agency to make or materially inform decisions affecting individuals.
Legislate independent audit and pre-deployment assessment for high-risk AI, following the UK model of binding codes of practice issued by existing regulators rather than a single consolidated AI Act, preserving sector expertise while closing the enforcement gap.
Mandatory data residency should apply to any AI system used by a Commonwealth agency to make or materially inform decisions about individuals, including through third-party providers, with limited and published exemptions on national security grounds.
A defined share of compute capacity built under the Australian Standards for AI should be reserved on published terms for Australian government, research and public-interest use, including narrow domain-specific models trained on Australian data in health, disability services, agriculture and resources.
Australia should pursue formal alignment with the United Kingdom and comparable jurisdictions on minimum mandatory requirements for advanced AI assessment, audit and data residency, with a view to a common standard rather than divergent national regimes.