From the record

What 170 submitters told the committee

The pipeline sorted the submitters into 23 topics and 7 themes, and wrote a claim and a description for each. This page shows those claims with the figures behind them. Every number is computed when the page loads.

  1. 01National AI scalingtie wider use to proven value, clear accountability and domestic capability3 topics · 170 submitters
  2. 02National AI adoptionretain control of critical dependencies while scaling proven uses8 topics · 165 submitters
  3. 03Small and regional business AI usefund tailored support beyond basic trials4 topics · 158 submitters
  4. 04Organisation-wide AI adoptionReplace isolated trials with staged, accountable implementation2 topics · 147 submitters
  5. 05Creative, cultural and identity rightsmake consent, payment and provenance enforceable2 topics · 131 submitters
  6. 06AI-enabled abuse and exploitationimpose preventive duties to protect vulnerable people and identity2 topics · 127 submitters
  7. 07Data-centre growthrequire operators to manage energy, water and community impacts2 topics · 23 submitters
The topics in the galaxy · size is reach, warmth is how contested
Topics, by how many submitters engage them
169
169
163
158
151
146
146
129
127
125
13 more topics, all in the reading below
  • Supports
  • Builds on
  • Mixed
  • Redirects
  • Opposes
  • Unclear
Theme 013 topics170 submitters135 positions

National AI scaling

tie wider use to proven value, clear accountability and domestic capability

The sources argue that Australia should expand AI use only where it delivers measurable public or economic value, with responsibility traceable from development and procurement through deployment, review and redress. They connect successful scaling to human judgement, workforce and regional support, continuous assurance, and targeted Australian capability. They differ on whether the priority should be lifecycle duties, operational controls, institution-led implementation, government procurement or sovereign capacity in high-consequence areas.

Positions
86% supportive · 3 contesting · 16 unclear
Sector
Government 30% · Company 29% +4

Topics in this theme 3

  1. Trace responsibility across AI supply chains

    The proposition is that risk-based governance can enable beneficial adoption only if responsibility remains clear across the AI lifecycle and supply chain. It seeks targeted closure of gaps in existing law, workable compliance, lifecycle assurance for high-risk systems, stronger regulatory capacity and better evidence about emerging harms. It differs from operational-assurance and automated-decision clusters by focusing on general allocation of responsibility among developers, deployers and other supply-chain actors.

    169 submitters
  2. Unite deployment, sovereignty and trust

    The cluster argues for moving from experimentation to organisation-wide AI use that produces measurable economic and service benefits. It calls for proportionate governance, accountable institutions, human judgement, challenge rights, Australian skills and infrastructure, sovereign capability and protection of information integrity. Compared with other broad adoption clusters, it most explicitly treats deployment, domestic value retention and public trust as an integrated national scaling agenda.

    169 submitters
  3. Scale governed, reviewable AI uses

    The proposition is that AI should be scaled where it demonstrates public or economic value, supported by shared infrastructure and sovereign capability without removing agency responsibility or user choice. It seeks adaptable risk-based governance, stronger controls for high-impact uses, transparency, human authority, review rights and fair workforce participation. Unlike the broader scaled-adoption cluster, its distinguishing test is demonstrable value combined with continued agency-level accountability.

    158 submitters
Read the themekey points and quotes · who says what · submitters that push back
Theme 028 topics165 submitters167 positions

National AI adoption

retain control of critical dependencies while scaling proven uses

The sources argue that Australia should expand AI uses that show clear public or economic value while retaining practical control over critical infrastructure, data and expertise. They link this selective sovereignty to clear responsibility, continuous assurance, human review and workable rules across AI supply chains and operations. Some focus on resilience against foreign dependence and Australian authority over data, while others stress agency accountability, commercially durable capability or tailored support for regional businesses.

Positions
93% supportive · 9 contesting · 3 unclear
Sector
Company 39% · Civil Society 21% · Industry Group 14% +3
Most-cited
+28
Topics, by how many submitters engage them
163
146
129
114
64
37
34
2
The same topics in the galaxy · size is reach, warmth is how contested

Topics in this theme 8

  1. Target durable high-consequence capacity

    The proposition is that Australia should concentrate sovereign capability where AI consequences and external dependencies are greatest. It calls for domestic infrastructure, expertise and commercially sustainable capacity, with specialised models built around Australian data and sector strengths, continuous assurance and identifiable accountability. Compared with the adjacent sovereignty cluster, it is distinguished by its emphasis on commercial durability and ongoing evidence-based assurance for high-consequence capability.

    163 submitters
  2. Govern complete AI workflows continuously

    The cluster argues that risk-based rules become effective only when assurance is continuous and examines complete systems, workflows and transitions from AI advice to action. It seeks traceable authority, transparency, review and redress, along with procurement standards, common evidence and regulatory coordination. Unlike general lifecycle governance, its distinctive concern is translating policy into operational controls over how interconnected AI-enabled workflows function in practice.

    146 submitters
  3. Govern clinical use across lifecycles

    The proposition is that evidence-based healthcare AI can improve care, access, productivity and research if governed according to its clinical function and risk. It calls for nationally consistent health-specific rules, lifecycle management, distributed accountability, meaningful clinical judgement and workforce capability. Unlike general lifecycle-governance clusters, it grounds assurance and responsibility in clinical use, patient safety and the organisation of healthcare delivery.

    129 submitters
  4. Prioritise control, resilience and data authority

    The cluster argues that sovereignty should mean practical control across the AI system rather than complete technological self-sufficiency. It seeks selective sovereign controls for critical uses, resilient domestic infrastructure, alternatives to foreign dependence, specialised models based on Australian strengths and consent-based governance of Australian data. Unlike the other sovereignty cluster, it places greater emphasis on community authority, data provenance and strategic resilience against external dependency.

    114 submitters
  5. Formalise councils in AI policy

    The proposition is that local government should become a formal partner in national AI policy rather than an end-user expected to follow a uniform central model. It calls for shared capability, coordinated procurement, service redesign, embedded safeguards and evaluation while preserving local human expertise and accountability. It differs from general public-sector adoption clusters by centring councils’ institutional role, diverse needs and purchasing power.

    64 submitters
  6. Protect judgement through AI design

    The proposition is that preserving human cognitive capability should be an explicit objective of AI policy. It advocates pro-cognitive system design, educational safeguards and meaningful oversight to counter unreliable outputs and uncritical reliance while retaining beneficial uses. Unlike broader human-in-the-loop governance clusters, it focuses on the long-term erosion of judgement and its consequences for education, workplaces, democracy and national security.

    37 submitters
  7. Bind government to reviewable limits

    The proposition is that government use of automated decision-making requires a binding whole-of-government framework. It seeks democratic and risk-based limits on automation, visible and explainable decisions, review rights, system-wide transparency, independent oversight and genuine human judgement backed by public accountability. It differs from general high-risk AI governance by focusing specifically on the legality, contestability and accountability of public administrative decisions.

    34 submitters
  8. Manage transition, access and safety

    The cluster argues that autonomous passenger transport should be governed as a phased social and economic transition, not merely deployed as a new technology. It seeks worker protections, equitable geographic and accessible services, municipal congestion-management powers, public-transport integration and safety arrangements for rare or ambiguous failures. It differs from general AI safety and workforce clusters by concentrating these requirements on passenger mobility systems and urban transport markets.

    2 submitters
Read the themekey points and quotes · who says what · submitters that push back
Theme 034 topics158 submitters66 positions

Small and regional business AI use

fund tailored support beyond basic trials

The sources argue that success should be measured by productive, organisation-wide AI use, not simple uptake, and that smaller businesses need practical support to move beyond experimentation. One stresses trusted, end-to-end implementation services and workforce redesign; the other focuses on regional constraints, data readiness and government procurement as a demonstration tool.

Positions
86% supportive · 6 contesting · 3 unclear
Sector
Company 52% · Industry Group 24% · Academia 8% +3
Most-cited
+12
Topics, by how many submitters engage them
151
88
58
54
The same topics in the galaxy · size is reach, warmth is how contested

Topics in this theme 4

  1. Tailor productive adoption support

    The proposition is that policy should promote productive AI use rather than adoption for its own sake, with particular support for small and regional businesses that have limited implementation capacity. It calls for tailored assistance, data readiness, workforce capability, manageable risk-based regulation and government leadership through procurement and governed deployment. Unlike the other small-business cluster, it specifically highlights regional constraints and the demonstration role of government procurement.

    151 submitters
  2. Build portable skills and better work

    The cluster argues that AI is more likely to reshape tasks and skill requirements than eliminate entire occupations, making job quality and transition design decisive. It calls for lifelong and portable practical skills, worker participation, human oversight, protection of entry pathways and equitable access to training and opportunity. Unlike business-adoption clusters, it centres the worker-level distribution of transition risks and the redesign of work as prerequisites for productivity.

    88 submitters
  3. Fund trusted end-to-end adoption support

    The cluster argues that policy should measure deep and productive AI use rather than headline uptake, particularly among smaller businesses stuck at basic experimentation. It seeks practical end-to-end assistance delivered through trusted channels, workforce and work-design support, and understandable risk-based compliance. Unlike the regional-business cluster, it most strongly emphasises integrated implementation services and durable capability as the route from basic use to organisation-wide adoption.

    58 submitters
  4. Build domestic robotics adoption pathways

    The cluster argues that robotics and physical AI should be recognised as a distinct national capability requiring support beyond general software adoption. It seeks end-to-end funding from investigation through sustained deployment, lower adoption costs for smaller firms, specialist skills, worker participation and stronger domestic suppliers and markets. Its distinguishing feature is the combined emphasis on embodied systems, industrial deployment and an Australian robotics supply chain.

    54 submitters
Read the themekey points and quotes · who says what · submitters that push back
Theme 042 topics147 submitters39 positions

Organisation-wide AI adoption

Replace isolated trials with staged, accountable implementation

The sources argue that organisations should move beyond fragmented AI trials towards coordinated implementation with clear responsibility, human judgement and workforce preparation. Some call for a repeatable institutional model with independent evaluation and measurable public value. Others offer more provisional support, stressing the need to balance innovation with safety, rights and integrity safeguards.

Positions
33% supportive · 0 contesting · 26 unclear
Sector
Civil Society 23% · Industry Group 21% +4

Topics in this theme 2

  1. Embed staged, evaluated implementation

    The cluster argues that successful adoption should be built through existing institutions, human judgement and clearly allocated responsibility. It calls for workforce capability as core infrastructure, staged and reusable implementation, independent evaluation and coherent operational governance tied to measurable public value. Compared with the more provisional organisational-adoption cluster, this cluster presents a firmer institutional model centred on repeatability, evaluation and accountable implementation.

    146 submitters
  2. Coordinate trials with workforce safeguards

    The proposition is that organisations should replace fragmented AI trials with coordinated and responsible implementation while developing workforce and education capability. It points toward balancing innovation with safety, rights, integrity and public-interest protections. Unlike neighbouring implementation clusters, its conclusions are expressly provisional because the underlying material appears to provide supporting evidence rather than a settled policy platform.

    62 submitters
Read the themekey points and quotes · who says what · submitters that push back
Theme 052 topics131 submitters51 positions

Creative, cultural and identity rights

make consent, payment and provenance enforceable

The sources argue that people need practical control over AI uses of creative works, cultural knowledge, identity, performance and likeness, backed by transparency, payment and effective remedies. One focuses on copyright licensing, training-data records and provenance, while the other extends protection to Indigenous-led cultural governance and unauthorised digital replicas.

Positions
90% supportive · 3 contesting · 2 unclear
Sector
Civil Society 27% · Industry Group 24% +4

Topics in this theme 2

  1. Enforce consent, payment and identity control

    The cluster argues that enforceable accountability must extend across the AI lifecycle to protect creative material, cultural knowledge, personal identity, performance and likeness. It seeks transparent and remunerated use, practical mechanisms for exercising rights, Indigenous-led governance of cultural knowledge and data, and coordinated enforcement. Unlike the copyright-specific cluster, it advances a wider rights infrastructure covering collective cultural authority and unauthorised digital replication as well as creative remuneration.

    125 submitters
  2. Require licensing, provenance and remedies

    The proposition is that creator consent, control and payment should remain central while lawful AI innovation continues. It calls for strong and clarified copyright protection across the AI lifecycle, transparent licensing markets, auditable training-data records, provenance and machine-readable rights information, clear responsibility and affordable Australian remedies. Unlike the broader rights-infrastructure cluster, it focuses specifically on copyright licensing and the evidence needed to make creator rights enforceable.

    110 submitters
Read the themekey points and quotes · who says what · submitters that push back
Theme 062 topics127 submitters42 positions

AI-enabled abuse and exploitation

impose preventive duties to protect vulnerable people and identity

The sources argue that AI providers should be required to prevent foreseeable abuse and protect people’s safety, consent and control over their identity. Some focus on synthetic media and harms to children and other vulnerable people, while others address high-risk AI uses more broadly. A further strand seeks enforceable control over likeness, creative work and cultural knowledge, including payment and Indigenous-led governance.

Positions
93% supportive · 2 contesting · 1 unclear
Sector
Civil Society 38% · Academia 33% +3

Topics in this theme 2

  1. Treat synthetic harms as systemic

    The cluster argues that AI-enabled abuse and synthetic content are foreseeable systemic harms for which providers must bear preventive responsibility. It calls for specific protections for children and other people at heightened risk, rights protection in sensitive services and consequential decisions, coordinated regulators and stronger domestic technical capability. Unlike the adjacent vulnerable-person cluster, its defining focus is abuse enabled or amplified by synthetic media and related AI capabilities.

    127 submitters
  2. Impose proactive provider safety duties

    The proposition is that preventing harm to children and vulnerable people should be a central objective of AI governance. It seeks proactive duties on providers, targeted safeguards and testing for high-risk applications, effective enforcement, coordinated technology-neutral reform and technically capable oversight. Unlike the synthetic-abuse cluster, it is distinguished by its broader provider-duty framework for preventing harms across high-risk AI applications before they occur.

    84 submitters
Read the themekey points and quotes · who says what · submitters that push back
Theme 072 topics23 submitters10 positions

Data-centre growth

require operators to manage energy, water and community impacts

The sources argue that expanding data centres should be governed by consistent requirements for clean electricity, efficient resource use, transparent reporting and suitable siting. One focuses on water use, environmental standards and benefits for host communities. The other stresses grid reliability, demand flexibility and making operators bear the infrastructure costs their growth creates.

Positions
90% supportive · 0 contesting · 1 unclear
Sector
Civil Society 50% · Government 30% +1

Topics in this theme 2

  1. Mandate clean, efficient, community-benefiting growth

    The proposition is that data-centre expansion requires nationally consistent and mandatory environmental standards. It calls for additional clean electricity, efficient operation, transparent energy and water reporting, water-demand reduction, suitable siting and tangible benefits for host communities. Unlike the other data-centre cluster, this one is distinguished by its broader sustainability and social-licence agenda, especially water use and community benefit-sharing.

    23 submitters
  2. Make operators manage grid impacts

    The cluster argues that data-centre growth must be integrated with electricity-system, climate, infrastructure and economic planning. It seeks additional renewable supply funded by operators, allocation of infrastructure costs to those operators, demand flexibility, prudent capacity planning, better siting and mandatory efficiency and resource reporting. Unlike the sustainability-focused data-centre cluster, its central concern is preventing uncertain demand from causing grid overbuilding, reliability problems or shifted infrastructure costs.

    12 submitters
Read the themekey points and quotes · who says what · submitters that push back