Submission 365
Mr Noah Quinlan
Submitter group: Individuals.
Position labels are not shown for individual submitters.
An AI model identified the submitter group and wrote the summary, key points and recommendation summaries. Counts do not measure importance or merit.
Summary
Read the original submission for the full text.
Argues for risk-based AI rules, individual freedoms, accountable government use, sovereign capability and public value. Recommends that large facilities pay infrastructure costs and meet environmental assessment, reporting, lifecycle and verification requirements. Supports development with safeguards; the attachment rejects established electromagnetic harm. Declares People First membership but submits individually, not for the party.
Key points
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.
Argues that Australia should welcome useful technology while retaining the ability to decide how, where and on what terms it is used.
Australia should welcome useful technology while retaining the ability to decide how, where and on what terms it is used.
Says existing privacy, consumer, administrative, criminal, workplace, discrimination and copyright law should be used where it works, with targeted reform where AI creates a genuine gap.
Existing privacy, consumer, administrative, criminal, workplace, discrimination and copyright law should be used where it works, with targeted reform where AI creates a genuine gap.
Argues sovereign AI capability requires practical control over strategically important compute, datasets, skills, intellectual property and supply chains, beyond servers on Australian soil.
Sovereign AI capability should mean more than servers sitting on Australian soil. Australia should retain practical control over strategically important compute, datasets, skills, intellectual property and supply chains.
Recommends that AI regulation preserve freedom of speech, privacy and due process, not create a general government power to score citizens' behaviour, and require explicit legislative authority for high-impact surveillance.
Accordingly, I recommend that AI regulation should preserve freedom of speech and association, privacy and due process; should not create a general power for government to infer or score citizens’ behaviour; and should require explicit legislative authority for high-impact uses of AI surveillance or biometric identification.
Argues that the case for reform is strongest where government exercises public power, and that voluntary settings should not be the final safeguard for high-impact decisions.
The case for reform is strongest where government exercises public power; voluntary settings should not be the final safeguard for high-impact decisions.
Argues that the costs created by very large AI facilities should not be shifted to households, existing businesses or the public.
I support the underlying principle that the costs created by very large AI facilities should not simply be shifted to households, existing businesses or the public balance sheet.
Recommendations
Recommendations are shown beside the submission’s own words. Listed in the same order as on its page in the interactive guide.
Adopt a risk-based national AI framework with enforceable baseline duties (accountability, privacy, security, data governance, testing, transparency, human oversight, contestability), with proportionately stronger obligations for high-impact and critical systems.
Adopt a risk-based national AI framework with enforceable baseline duties for accountability, privacy, security, data governance, testing, transparency, human oversight and contestability, with proportionately stronger obligations for high- impact and critical systems.
Use proportionate baseline duties for low-risk AI; require documented assessment/testing/governance/complaints for medium risk, stronger assurance/oversight/records/enforceable rights for high-impact systems, and extra resilience/supply-chain controls for critical systems.
The preferred model is risk-based: low-risk uses receive proportionate baseline duties; medium-risk systems require documented assessment, testing, governance and complaints mechanisms; high-impact systems require stronger assurance, human oversight, record keeping and enforceable rights; and critical systems require additional resilience and supply-chain controls.
Prohibit unjustified AI-enabled digital coercion, generalised behavioural scoring and indiscriminate surveillance, with explicit legislative authority required for high-impact biometric or surveillance uses.
Protect individual freedom by prohibiting unjustified AI-enabled digital coercion, generalised behavioural scoring and indiscriminate surveillance, with explicit legislative authority required for high-impact biometric or surveillance uses.
Preserve freedom of speech and association, privacy and due process in AI regulation.
Accordingly, I recommend that AI regulation should preserve freedom of speech and association, privacy and due process; should not create a general power for government to infer or score citizens’ behaviour; and should require explicit legislative authority for high-impact uses of AI surveillance or biometric identification.
For material Commonwealth AI, disclose purpose and whether it makes, recommends or materially influences individual decisions, and name an accountable senior official for each system.
For Commonwealth agencies using material AI systems, I recommend minimum safeguards including: public identification of material AI use and its purpose, including whether the system makes, recommends or materially influences decisions about individuals; a named accountable senior official for each material AI system;
For Commonwealth AI, provide meaningful human review of significant benefits, penalties, eligibility, employment, immigration, legal-status and essential-service decisions; enable challenge/correction/review, reconstructable decision records and independent high-impact assurance.
meaningful human review for significant decisions affecting benefits, penalties, eligibility, employment, immigration, legal status or access to essential services; a genuine avenue to challenge, correct and review an AI-assisted decision; and record-keeping sufficient to reconstruct the basis of a material decision, with independent assurance for high-impact systems.
Establish a sovereign-AI capability program covering domestic compute, skills, research, model development, secure Australian datasets and interoperability, emphasising domain-specific capability with clear strategic value.
Establish a sovereign-AI capability program covering domestic compute, skills, research, model development, secure Australian datasets and interoperability, with emphasis on domain-specific capability where it provides clear strategic value.
Assess publicly supported projects for dependency and lock-in; support Australian-owned/controlled capability where justified, interoperable systems, provider portability and secure Australian datasets.
Publicly supported projects should therefore be assessed for dependency and lock-in, not simply location. The Commonwealth should support Australian-owned or Australian-controlled capability where justified, interoperable systems, portability between providers and secure Australian datasets.
Use Commonwealth procurement to expand opportunities for capable Australian AI businesses and avoid unnecessary vendor lock-in.
Use Commonwealth procurement to expand opportunities for capable Australian AI businesses and avoid unnecessary vendor lock-in
For major Commonwealth-supported AI and infrastructure projects, include measurable Australian training, apprenticeships and skills outcomes.
Major Commonwealth-supported AI projects and major AI infrastructure projects should include measurable Australian training, apprenticeships and skills-development outcomes.
Keep SME and family-business compliance proportionate, with plain-English guidance and standardised low-risk pathways.
Keep SME and family-business compliance proportionate, with plain-English guidance and standardised low-risk pathways.
Require high-impact uses to have clear data provenance, documented authority, secondary-use limits, security controls and correction avenues.
Information that can technically be collected should not thereby become available for model training, profiling, inference or behavioural manipulation without a clear lawful basis and appropriate safeguards. High-impact uses should require clear data provenance, documented authority, limits on secondary use, security controls and avenues for correction.
Protect Australian copyright and cultural content through provenance requirements and clear, enforceable rights or lawful licensing arrangements for training uses.
Protect Australian copyright and cultural content through provenance requirements and clear, enforceable rights or lawful licensing arrangements for training uses.
Prioritise practical harms (scams, impersonation, deepfakes, sextortion, child exploitation) through targeted measures rather than broad censorship or pervasive surveillance.
Prioritise practical harms—scams, impersonation, deepfakes, sextortion and child exploitation—through targeted measures rather than broad censorship or pervasive surveillance.
Use clear provenance, traceability where appropriate, rapid reporting pathways and strong penalties for deliberate AI misuse.
For harms, the Committee should prioritise practical problems already capable of causing direct damage: fraud and scams, impersonation, deepfakes, extortion, exploitation of children and other vulnerable people, and malicious automated attacks. Broad censorship powers or pervasive surveillance are not an acceptable substitute for targeted enforcement. Clear provenance, traceability where appropriate, rapid reporting pathways and strong penalties for deliberate misuse are preferable.
For government and critical-infrastructure AI, require dependency mapping, secure updates, incident reporting, supply-chain assurance, tested fallbacks and safe operation when foreign providers or critical components are unavailable.
Systems used by government and critical infrastructure should therefore be subject to stronger requirements than ordinary consumer applications. Those requirements should include dependency mapping, secure update mechanisms, incident reporting, supply-chain assurance, tested fallback arrangements and the ability to operate safely if a foreign provider or critical component becomes unavailable.
Require large AI facilities to bear the full incremental cost of the power, water, network and other infrastructure they require, and consider requiring own additional power and water capacity where practicable.
Require large AI facilities to bear the full incremental cost of the power, water, network and other enabling infrastructure they require, and consider the People First principle that such facilities should provide their own additional power and water capacity where practicable.
Include incident response in cyber and supply-chain resilience requirements for government and critical-infrastructure AI.
Require strong cyber and supply-chain resilience for AI used by government and critical infrastructure, including dependency mapping, incident response, secure updates and tested fallback arrangements.
Make material public AI support conditional, where proportionate, on measurable Australian R&D, training, procurement, compute access, infrastructure contributions, transparent corporate structures and appropriate tax compliance, while welcoming beneficial foreign investment.
The Commonwealth should therefore make material public support for AI conditional, where proportionate, on measurable Australian benefits such as local R&D, training, domestic procurement, access to compute for Australian innovators, contribution to enabling infrastructure, transparent corporate structures and appropriate tax compliance. This is not a proposal to exclude foreign investment. It is a proposal to make foreign investment serve clearly demonstrated Australian interests.
Spread regional AI benefits through research partnerships, shared compute, skills and applications in agriculture, resources, manufacturing, health and regional services, without disproportionate infrastructure/environmental costs.
Develop a practical framework to spread AI benefits to regions through research partnerships, shared compute, skills and applications in agriculture, resources, manufacturing, health and regional services, without transferring disproportionate infrastructure or environmental costs to regional communities.
Make regional deployment conditional on adequate electricity, water, telecommunications and environmental capacity, without presuming regional communities should host resource-intensive facilities.
The Commonwealth should support regional research partnerships, secure shared compute for smaller firms and researchers, and practical AI applications relevant to regional industries and public services. Regional deployment should, however, remain conditional on adequate electricity, water, telecommunications and environmental capacity rather than creating a presumption that regional communities should host resource-intensive facilities.
Review the framework periodically so regulatory settings can respond to changes in AI capability without unnecessary duplication.
Review the framework periodically so that regulatory settings can respond to changes in AI capability without unnecessary duplication.
Examine concentration risk, so the digital economy does not depend on a small number of overseas platforms, models, cloud providers, semiconductor suppliers or software dependencies for essential services.
Australia should also examine concentration risk. A resilient digital economy cannot depend on a small number of overseas platforms, models, cloud providers, semiconductor suppliers or software dependencies for essential services.
Ask the Committee to read this submission with the separate Senate submission, take account of the Tasmanian petition tabled on 2 September 2026, and take note of how Queensland responds to the submitter's petition.
I respectfully ask the Committee to consider these recommendations, to read this submission together with my separate submission to the Senate inquiry into artificial intelligence and data centres, to take account of the Tasmanian petition tabled on 2 September 2026, and to take note of how the Queensland Parliament and Government respond to my petition.
Prioritise practical AI applications in agriculture, resources, health/medical research, advanced manufacturing, financial services, emergency management and defence.
Australia should prioritise practical applications in agriculture, resources, health and medical research, advanced manufacturing, financial services, emergency management and defence, consistent with the Committee’s Terms of Reference.
Ensure the national AI framework addresses physical infrastructure, resource allocation, community consultation and public value rather than leaving these to fragmented state/local processes.
They strengthen the case for the Commonwealth to ensure that the national AI policy framework addresses the physical infrastructure, resource allocation, community consultation and public-value questions that can otherwise be left to fragmented state and local processes.
Establish a national framework for cumulative environmental assessment of large AI data centres and their associated infrastructure, including where multiple facilities interact within a region.
Establish a national framework for cumulative environmental assessment of large AI data centres and their associated infrastructure, including where multiple facilities may interact within a region.
Assess cumulative regional effects, where relevant, across electricity/water infrastructure, land clearing/habitat, threatened species, noise/light, backup emissions, construction, waste, stormwater/contamination and interactions between facilities before approving concentrations of large facilities.
Cumulative assessment should consider, where relevant: - Electricity generation, transmission and distribution infrastructure; - Water supply and wastewater infrastructure; - Land clearing and habitat fragmentation; - Threatened species and ecological communities; - Noise and lighting; - Air emissions from backup generation; - Construction impacts; - Waste and equipment disposal; - Stormwater and land contamination risks; and - Interactions between multiple facilities in the same region. The objective should not necessarily be to prevent development. It should be to ensure that governments understand cumulative environmental and infrastructure consequences before approving or facilitating concentrations of very large facilities.
Require site-selection assessment of suitable developed land, electricity/water capacity, sensitive receptors, biodiversity/threatened species/corridors, habitat fragmentation/clearing, hazards, off-site infrastructure and cumulative regional effects; avoid sensitive sites wherever reasonably practicable.
I recommend that large AI infrastructure proposals be required to demonstrate that site selection has considered: - Existing industrial or otherwise suitable developed land; - Availability and capacity of electricity and water infrastructure; - Proximity to homes, schools and other sensitive receptors; - Biodiversity, threatened species and ecological corridors; - Habitat fragmentation and land-clearing requirements; - Flood, bushfire and other relevant natural hazards; - The infrastructure required outside the site boundary; and - The cumulative effects of other proposed facilities in the surrounding region. Environmentally sensitive sites should be avoided wherever reasonably practicable rather than relying exclusively on mitigation or offsetting after a site has been selected.
For facilities above an appropriate capacity threshold, require at least annual public reporting of electricity use/peak demand/contracted and actual supply, water use/source/quality, cooling/efficiency, backup capacity/fuel, incidents, retired equipment/outcomes, refrigerants/losses, monitoring and capacity changes.
For facilities above an appropriate capacity threshold, operators should publicly report, at least annually: - Electricity consumption; - Peak electricity demand; - Contracted and actual electricity supply; - Water consumption; - The source and quality of water used; - Cooling technology; - Power Usage Effectiveness and other appropriate efficiency measures; - Backup-generation capacity and fuel consumption; - Significant environmental incidents; - Material quantities of electronic equipment retired; - Reuse, recycling and disposal outcomes; - Refrigerants used and significant refrigerant losses; - Relevant environmental monitoring results; and - Material changes to the facility's approved operating capacity.
Require transparent reporting of development status, distinguishing speculative connection requests, proposed capacity, approved or contracted capacity, construction and actual operational demand.
Require transparent reporting of development status, distinguishing speculative connection requests, proposed capacity, approved or contracted capacity, construction and actual operational demand.
Do not treat headline project pipelines as inevitable built capacity or consumption in public policy.
Public policy should therefore not rely on headline project pipelines as though all proposed capacity will necessarily be constructed and consumed.
Set national minimum requirements or guidance on operational/low-frequency/tonal/night noise, light spill, backup-generator emissions/testing frequency/duration, fuel spills, construction and cumulative local impacts; make independent monitoring verification available for materially community-impacting facilities.
I recommend national minimum requirements or guidance for large AI facilities covering: - Operational noise; - Low-frequency and tonal noise where relevant; - Night-time noise impacts; - Lighting and light spill; - Emissions from backup generators; - Generator testing frequency and duration; - Fuel storage and spill risks; - Construction noise and environmental impacts; and - Cumulative impacts where multiple facilities operate in proximity. Where a facility is sufficiently large to create material community impacts, independent verification of monitoring data should be available.
Include stormwater in national minimum expectations for local environmental impacts of large AI facilities.
Develop national minimum expectations for local environmental impacts, including noise, low-frequency/tonal noise where relevant, lighting, backup-generator emissions, construction impacts and stormwater.
Require appropriate electromagnetic-field assessment for major high-voltage infrastructure associated with large data centres where sensitive receptors may be affected, while recognising current evidence does not establish ordinary ELF or RF exposure as a demonstrated hazard.
Require appropriate electromagnetic-field assessment for major high-voltage infrastructure associated with large data centres where sensitive receptors may be affected, while recognising that current evidence does not establish ordinary ELF or RF exposure from such infrastructure as a demonstrated health or ecological hazard.
Where major data-centre electrical infrastructure is near sensitive receptors, comply with applicable Australian radiation-protection standards/guidance, publish relevant field assessments and continue environmental-effects research.
Rather, where a large data-centre development requires new transmission lines, substations or other major electrical infrastructure near homes, schools or other sensitive receptors, I recommend: - Assessment of electromagnetic-field exposure as part of site and infrastructure planning; - Compliance with applicable Australian radiation-protection standards and guidance; - Public availability of relevant field assessments where major infrastructure is proposed near sensitive receptors; and - Continued research into possible environmental effects of anthropogenic electromagnetic fields.
Support continued independent research into electromagnetic-field effects on plants and animals, particularly where evidence gaps remain concerning newer technologies and natural environments.
Support continued independent research into electromagnetic-field effects on plants and animals, particularly where evidence gaps remain concerning newer technologies and natural environments.
Require large facilities to disclose annual/peak water use, source, potable/recycled/non-potable proportions, cooling technology, cooling energy and environmental trade-offs; demonstrate a reasonable outcome considering local water security and electricity-system energy use.
I therefore recommend that large facilities disclose: - Annual and peak water consumption; - Water source; - Potable, recycled and non-potable water proportions; - Cooling technology; - Energy consumption associated with cooling; and - Material environmental trade-offs considered when selecting the cooling system. Large facilities should demonstrate that their selected cooling approach represents a reasonable environmental outcome having regard to both water security and energy use in the relevant electricity system.
Require large AI data-centre equipment/e-waste plans covering replacement, reuse/refurbishment, valuable-material recovery, sound recycling, batteries/cooling equipment, hazardous components and final destinations; report retired quantities and reused/refurbished/recycled proportions.
I recommend that large AI data centres maintain equipment lifecycle and e-waste management plans covering: - Expected equipment replacement; - Reuse and refurbishment; - Recovery of valuable materials; - Environmentally sound recycling; - Disposal of batteries and cooling equipment; - Management of hazardous components; and - Final destinations of retired equipment. Operators should report quantities of major equipment retired and the proportion reused, refurbished or recycled.
Disclose major-cooling refrigerants, environmental characteristics and leak detection/prevention; explain technology choices in assessment where lower-impact alternatives are technically and economically practicable.
Large data centres should nevertheless disclose the refrigerants used in major cooling systems, their relevant environmental characteristics, and measures used to detect and prevent leakage. This would complement rather than duplicate existing Commonwealth regulation. Where lower-impact cooling or refrigerant technologies are technically and economically practicable, developers should explain their technology choice as part of environmental assessment.
Establish independent verification or auditing of environmental reporting for facilities above an appropriate capacity or environmental-risk threshold.
Establish independent verification or auditing of environmental reporting for facilities above an appropriate capacity or environmental-risk threshold.
Scale independent audits to facility size and environmental risk; these could verify energy/water, cooling, generators, monitoring, equipment/e-waste, refrigerants and environmental-condition compliance; impose enforceable consequences for material non-compliance.
I recommend that the Commonwealth establish independent verification requirements for large AI data centres. Depending on facility size and environmental risk, independent audits could verify: - Electricity and water consumption; - Cooling efficiency; - Generator operation; - Environmental monitoring; - E-waste and equipment-recycling records; - Refrigerant management; and - Compliance with environmental conditions. Material non-compliance should have enforceable consequences rather than relying exclusively on voluntary disclosure.
Coordinate with states/territories through existing ministerial/regulatory mechanisms on baseline assessment, siting, resource reporting, noise/local impacts, lifecycle and verification; retain stronger local requirements where water, biodiversity, infrastructure or community conditions warrant.
The Commonwealth should therefore work with states and territories through existing ministerial and regulatory mechanisms to establish a common national baseline for: - Environmental assessment; - Site selection; - Resource-use reporting; - Noise and local environmental impacts; - Equipment lifecycle management; and - Independent environmental verification. States and territories should retain the ability to impose stronger requirements where local water, biodiversity, infrastructure or community conditions warrant them.