Submission 299
Tech Council of Australia
Submitter group: Industry bodies.
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 Australia should be an AI maker, focusing on middleware, applications and domain-specific models rather than frontier general-purpose models. Recommends investment, copyright clarity with equitable outcomes protecting creators, adoption support, workforce and migration measures, targeted risk-based and internationally consistent regulation, and cyber security capability.
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 faces a choice between developing AI and being a purchaser and user of AI built elsewhere.
Australia faces a once-in-a-generation choice: whether to be a genuine global player in AI development and capture the socio-economic opportunity, or simply become purchasers and users of AI developed elsewhere and expect our government to manage its consequences.
Argues that the strongest economic opportunity lies in the middleware and application layers of the AI stack.
The strongest economic opportunity for Australia exists in layers 4 and 5 of the AI stack - the middleware & orchestration layer and the application & product layer - where Australian developers and entrepreneurs are best placed to compete globally and capture lasting economic value.
Argues that Australia’s Layer 3 opportunity lies in domain-specific models trained on high-quality Australian data in sectors of comparative advantage.
The more realistic and strategically sound opportunity at Layer 3 lies in developing narrower, domain-specific models trained on high-quality Australian data in sectors where Australia has a genuine comparative data and economic advantage, including but not limited to agriculture, mining, healthcare, legal services, and defence.
Argues that copyright clarity underpins Australia’s ambition to develop AI and host frontier AI development.
Copyright is therefore a legal precondition for AI development in Australia, and bears directly on Australia's ambition to be an AI builder, as well as a host for frontier AI development.
Argues that regulation should be based on the nature and scale of potential harm rather than the use of a particular technology.
Regulation should be based on the nature and scale of the potential harm, rather than the use of a particular technology.
Says current evidence does not show AI displacing workers or early-career roles to date.
Australians are genuinely concerned about AI and what it will mean for their jobs. However, current evidence does not show that AI is displacing workers or early-career roles to date.
Recommendations
Recommendations are shown beside the submission’s own words. Listed in the same order as on its page in the interactive guide.
Assess and publish measures of national AI capability by layer of the AI stack, including investment share, number of exporting Australian AI firms, AI-skilled workforce depth and accessible compute.
Assess and publish measures of national AI capability by layer- including the share of investment reaching different layers of the stack, the number of Australian firms exporting AI products, AI-skilled workforce depth, and accessible compute.
Use this evidence to position Australia where it can genuinely lead, through policies and support programs that encourage and accelerate local AI development.
Use this evidence to position Australia where it can genuinely lead to harness the economic and social benefits of AI, by developing policies and support programs that actively encourage and accelerate the development of AI locally.
Treat sovereign model capability as achievable and valuable, targeting specialised, open-weights-derived and multimodal models rather than frontier general-purpose models.
Government should therefore treat sovereign model capability as achievable and valuable, targeted at specialised, open-weights-derived and multimodal models rather than at frontier general-purpose models.
Implement the findings of the Strategic Examination of Research and Development’s Ambitious Australia report.
Australia's research and development settings are already the subject of comprehensive review through the Strategic Examination of Research and Development (SERD) which reported in March 2026. The SERD’s Ambitious Australia report set out a system-wide program of reform spanning foundational research, business incentives, investment and capital, and workforce capability. The report sets out a roadmap to build a dynamic innovative environment in Australia that would support AI companies at layers 3, 4 and 5 of the stack. The TCA supports the direction of the proposed reforms and encourages the government to implement the report's findings.
Ensure programs and policy settings, including tax settings (CGT and RDTI), support founders to start and grow AI companies in Australia, including pre-accelerators, accelerators, capital and networks. Provide founders with early-stage ‘just-in-time’ knowledge.
Ensure programs are in place and policy settings support founders to start and grow AI companies in Australia. Programs include pre-accelerators and accelerators, that help founders with ‘just-in-time’ knowledge at the early stages of company development, as well as access capital and networks that support scaling. Supportive policies must necessarily include tax settings (including CGT and RDTI).
Ensure policy settings support the venture capital industry by unlocking private investment to fund Australian AI companies from seed stage through to scale.
Ensure policy settings support the further development of the venture capital industry by unlocking private sector investment that will fund Australian AI companies from seed stage through to scale.
Identify, prioritise and support national-interest domain-specific models, principally cyber security, defence, health and medical, agriculture and food, energy and environment, and resources.
Identify and prioritise narrow, domain-specific models that are in the national interest (principally cyber security, defence, health and medical, agriculture and food, energy and environment, and resources) and encourage and support their development.
Publish shared national public datasets, assets and AI training/testing environments under government stewardship, with clear usage, privacy and security guidelines; datasets must be high-value, discoverable, well-governed and interoperable.
Establish and publish shared national public datasets, assets and environments for AI training and testing under government stewardship, with clear guidelines on data usage, privacy and security. Datasets must be high-value, discoverable, well-governed and interoperable.
Upgrade government APIs, data interoperability, secure sharing and shared reference architectures, moving the APS from bespoke systems to common foundations for safe AI deployment at scale.
The government should upgrade APIs, improve data interoperability, build secure data-sharing capability and develop shared reference architectures, and move the APS from bespoke systems to common foundations that enable safe AI deployment at scale.
Proceed with privacy reform to modernise consumer protections, clarify compliance requirements and improve international consistency, including on facial recognition, automated decision-making and consumer data use and consent.
Proceed with privacy reform to modernise consumer protections while clarifying compliance requirements and improving international consistency, including on facial recognition, automated decision-making and consumer data use and consent.
Establish a standing national digital infrastructure advisory function with industry and states and territories to coordinate planning and streamline approvals for AI infrastructure, digital connectivity and associated renewable energy.
Establish a standing national digital infrastructure advisory function with industry and state and territory governments, to coordinate strategic planning and streamline approvals for AI infrastructure, the digital connectivity infrastructure it depends on, and associated renewable energy.
Treat data centres as an essential foundation for sovereign capability in the AI stack, with planning and permitting systems for data centres and digital connectivity equipped accordingly.
Treat data centres as an essential foundation for establishing Australia’s sovereign capability in the AI stack. Planning and permitting systems for the data centres and underpinning digital connectivity infrastructure should be equipped accordingly.
Ensure copyright supports Australian AI training and development while delivering equitable outcomes protecting creators’ and creative industries’ rights.
Australia’s copyright system should serve Australia’s national interests, enabling AI training and development in Australia while delivering equitable outcomes that protect the rights of creators and the creative industries. Australians can benefit from both flourishing AI and creative industries.
Implement copyright measures and open-data reform together so national datasets carrying third-party intellectual property can be assembled.
At the data layer, national datasets cannot be assembled until the legal position on training is settled. Our data assets in cultural collections, archives, scientific literature and government-held material typically carry third-party intellectual property rights, which is why copyright measures and open data reform must be implemented together;
Clarify responsibility for copyright-infringing AI application outputs.
At this layer, clarity is also required regarding responsibility for copyright infringing outputs.
Expeditiously develop a solution delivering clarity on the copyright conditions to be met so AI models or applications can be developed and refined in Australia.
Expeditiously develop a solution which delivers clarity on the copyright conditions to be met and enables AI models or applications to be developed and refined in Australia.
Encourage businesses to establish mature responsible-AI practices and integrate AI deeply into operations.
It is critical that businesses of all sizes adopt AI capably and responsibly, and the government should encourage businesses to establish mature responsible AI practices. The government should encourage integration of AI deeply into business operations, because the productivity gains come from depth of use rather than trial: according to Tech Council research, lifting SME adoption to 53 per cent with broader use inside adopting firms would unlock $93 billion in productivity uplift.6
Uplift digital literacy and responsible AI awareness for the Australian people, SMEs and NGOs, including guidelines, tools and assurance frameworks to operationalise responsible AI.
Uplift digital literacy and responsible AI awareness for the Australian people, SMEs and NGOs, including guidelines, tools and assurance frameworks that help organisations operationalise responsible AI.
Work jointly with industry associations and chambers of commerce to drive AI diffusion.
Work jointly with industry associations and chambers of commerce to drive diffusion.
Pair fee-free TAFE with targeted employer incentives (wage subsidies, matched training grants or refundable tax credits) to de-risk AI experimentation for SMEs.
Pair fee-free TAFE with targeted employer incentives - wage subsidies for trainees, matched training grants or refundable tax credits - to de-risk AI experimentation for SMEs.
Act as lead adopter and lead customer by accelerating the digital government agenda, modernising legacy platforms and reforming procurement to create pathways for Australian businesses to scale.
Act as lead adopter and lead customer by accelerating the digital government agenda, modernising legacy platforms; and reforming procurement to create pathways for Australian businesses to scale.
Invest in community understanding of the benefits of responsible AI and partner with trusted stakeholders on AI literacy programs so all Australians can use AI with confidence.
Invest in building community understanding of the benefits of responsible AI development and use in Australia, and partner with trusted stakeholders to develop and deploy AI literacy programs to ensure that all Australians are able to use AI with confidence.
Report regularly and publicly on how AI is changing the Australian labour market so skilling and transition policy stays proportionate to observed effects and can respond quickly if effects change.
Report regularly and publicly on how AI is changing the Australian labour market so that skilling and transition policy stays proportionate to observed effects and can move quickly if they change.
Treat the specialist AI workforce as a distinct national capability with planning across training, retention and migration.
Treat the specialist AI workforce as a distinct national capability with planning across training, retention and migration, because engineers, researchers and technical founders who build AI form a relatively small, globally mobile cohort.
Orient AI training toward real-world application and structure placements in technical degrees and research training so students and research candidates work on companies’ production AI systems before entering the workforce.
Orient AI training toward real-world application and build structured industry placement into technical degrees and research training, so that students and research candidates work on production AI systems inside companies before they enter the workforce.
Prepare the education and training system to deliver AI upskilling at national scale, working with industry and unions and prioritising workers least likely to be reached by employer-led training.
Prepare and enable the education and training system to deliver AI upskilling at national scale. Work with industry and unions on training that builds durable capability and prioritise the workers least likely to be reached by employer-led training and skills development.
Ensure the skilled migration system is responsive to technological change, beginning with occupation classification.
Ensure the skilled migration system is responsive to technological change, beginning with occupation classification.
Recognise elite entrepreneurial talent alongside research excellence.
Recognise elite entrepreneurial talent alongside research excellence.
Enable founders and operators who have built globally competitive technology businesses to work in Australia.
Australia's talent pathways, including the National Innovation Visa, currently prioritise academic and research distinction more highly than a record of building and scaling companies. This is not fit-for-purpose. Founders and operators who have built globally competitive technology businesses are crucial for an AI builder economy, and should be enabled to work in Australia.
Provide expedited, predictable processing pathways for specialist AI and machine-learning talent where employers show the expertise is not readily available locally, with clear standards and accredited sponsors.
Provide expedited and predictable processing pathways for specialist AI and machine-learning talent where employers can demonstrate that the required expertise is not readily available locally, supported by clear processing standards and recognition of accredited sponsors.
Publish a national AI regulatory roadmap and gap analysis clarifying how existing laws apply, identifying genuine gaps, allocating regulator responsibilities and explaining when targeted legislation or standards may be needed.
Publish a national AI regulatory roadmap and gap analysis to clarify how existing laws apply, identify genuine gaps, allocate regulator responsibilities and explain when targeted legislation or standards may be required.
Base regulation on potential harm, considering seriousness and reversibility, affected people’s number and vulnerability, system autonomy and material influence on decisions about individuals.
Regulation should be based on the nature and scale of the potential harm, rather than the use of a particular technology. Relevant factors include the seriousness and reversibility of harm, the number and vulnerability of people affected, the autonomy of the system, and whether AI materially influences decisions about individuals, such as those below.
Keep the regulatory framework technology-neutral and responsive to emerging applications, avoiding default classification of every use as higher risk.
The regulatory framework should be technology neutral to account for emerging applications, without needing to unnecessarily redraw the boundaries and to avoid treating every use of them as a higher risk application by default.
Address higher-risk use cases in adaptable regulatory guidance, developed by sector-aware regulators in consultation with relevant industry, rather than in legislation.
Higher-risk use cases are better addressed in regulatory guidance (as opposed to the legislation itself) which can be updated relatively quickly to respond to emerging applications, including by the regulators who are cognisant of the risks in their own sectors working in consultation with the relevant industry.
Avoid zero risk-tolerance, blanket bans and moratoria in risk-based AI regulation.
Risk-based regulation should not equate to zero risk-tolerance, as this is unworkable. Blanket bans and moratoria should also be avoided as they would have unintended consequences for the economy and digital sector, and impede emerging businesses that could otherwise deliver net positive benefits.
Adopt risk-based, role-based regulation reflecting harm seriousness and context and developers’, providers’, deployers’ and substantial modifiers’ control, knowledge and scale; simplify compliance for low-risk uses.
Adopt risk-based and role-based regulation. Obligations should reflect the seriousness and context of potential harm and the control, knowledge and scale of developers, providers, deployers and substantial modifiers. Low-risk uses should have simplified compliance pathways.
Co-design the proposed National AI Standards with government, industry, technical experts, civil society and affected stakeholders.
The proposed National AI Standards should be co-designed by the government together with industry, technical experts, civil society and affected stakeholders to ensure that they’re fit for purpose, given the highly technical and fast moving nature of the technology.
Provide Australian recognition for equivalent testing, assurance and certification done in trusted jurisdictions by aligning local AI standards with recognised international standards.
Any local AI standards should also align them with recognised international standards so companies can have Australia recognise equivalent testing, assurance and certification done in trusted jurisdictions for regulatory coherence.
Co-design internationally consistent standards with industry, align with ISO, NIST, OECD and other frameworks, support mutual recognition and give appropriate regulatory recognition to demonstrated coherence.
Work with industry to co-design internationally consistent standards. Align Australian requirements with recognised ISO, NIST, OECD and other international frameworks, support mutual recognition and give appropriate regulatory recognition to demonstrated coherence.
Establish proportionate, role-differentiated, internationally aligned transparency along the AI supply chain as a feature of the AI standards framework.
Establish proportionate, role-differentiated, internationally aligned transparency along the AI supply chain as a feature of the AI standards framework.
Make AI transparency measures meaningful, technically feasible, commercially sensitive and tied to clear objectives such as safety, accountability or data provenance rather than generic disclosure requirements.
Transparency measures that form part of Australian AI governance should be meaningful in practice, technically feasible, commercially sensitive and tied to a clear policy objective such as safety, accountability or data provenance, rather than broad, generic disclosure requirements.
Tailor disclosure to its audience: detailed confidential disclosure to safety auditors or regulators, technical and contractual disclosure to deployers, and clear, widely comprehensible public disclosure.
Further, to whom the disclosure is made is formative; confidential disclosure to a safety auditor or regulator can be detailed, disclosure to a deployer is likely technical and contractual, and public-facing disclosure needs to be clear and widely comprehensible.
Increasingly authenticate agents’ identities and accountable operators, particularly when agents transact, access sensitive information or interact with critical systems.
As AI systems become capable of acting autonomously rather than merely generating content, authentication should increasingly address the identity and accountable operator of the agent itself, particularly where agents transact, access sensitive information or interact with critical systems.
Lead by example in government procurement and apply existing Commonwealth standards for its own AI use: accountable officials, use-case registers, impact assessments, staff training and risk-based governance.
Lead by example in government procurement and apply existing Commonwealth standards for its own AI use, accountable AI officials, use-case registers, impact assessments, staff training and risk-based governance.
Strengthen the Australian AI Safety Institute and develop a competitive assurance ecosystem providing independent testing, evaluation, audit and certification.
Build Australia’s AI safety and assurance capability by strengthening the Australian AI Safety Institute and develop a competitive assurance ecosystem capable of providing independent testing, evaluation, audit and certification.
Set credible AI assurance standards and accreditation arrangements, with industry technical expertise, tools and implementation experience, and clear safeguards for independence and conflicts of interest.
A competitive AI assurance market can turn a compliance regime into: first, a market-access asset for Australian companies, and second, an exportable service industry in its own right.22 This will be achieved if the government sets credible standards and accreditation arrangements while industry contributes technical expertise, testing tools and implementation experience, accompanied by clear safeguards for independence and conflicts of interest.
Clarify the roles and responsibilities of the National AI Centre, Australian AI Safety Institute and Office of AI, and designate a point of entry for industry guidance.
Clarify the respective roles and responsibilities of the National AI Centre, Australian AI Safety Institute and Office of AI, and designate a point of entry for industry guidance.
Operationalise regulatory sandboxes and safe harbour provisions for controlled experimentation, particularly for agentic and multi-agent systems, and embed technical collaboration into regulatory design and review.
Operationalise regulatory sandboxes and safe harbour provisions to enable controlled experimentation and structured industry collaboration, particularly for agentic and multi-agent systems, and embed technical collaboration into regulatory design and review.
Give regulatory sandboxes clear entry criteria, regulator supervision and mechanisms to share learning, including for agentic and multi-agent systems.
To encourage responsible AI development and help government and industry take a more forward-looking approach and understand new risks before rules are established, regulatory sandboxes can be established with safe harbour provisions. Such sandboxes are useful for emerging applications with clear entry criteria, regulator supervision, and a way to share what's learned, including for agentic and multi-agent systems.
Apply Australian safety arrangements cohesively to imported and remotely supplied AI services and support cooperation between Australian deployers and upstream providers.
Despite this, many systems used by Australians will continue to be developed overseas. Australian safety arrangements must therefore apply cohesively to imported and remotely supplied services, and support cooperation between Australian deployers and upstream providers.
Recognise and build on existing technical governance, including pre/post-deployment assessment, evaluation, red-teaming, audit, documentation, data governance, monitoring, incident response and shared vulnerability reporting.
Since AI industry technical governance is already substantial, regulation should recognise and build on existing practices to achieve this cohesiveness. Current systems include pre- and post-deployment Australian Government, Independent review of the Enhanced Regulatory Sandbox – Final report, Treasury, tabled 1 July 2026, ISBN 978-1-923278-53-0. Government response published 7 September 2026. Singapore’s AI Verify Foundation and the assurance consortium in the United Kingdom, for example. The UK Government estimates the size of its AI assurance market at £1.01 billion in 2024, rising to £18.8 billion by 2035; https://www.gov.uk/government/publications/assuring-a-responsible-future-for-ai and https://www.gov.uk/government/publications/trusted-third-party-ai-assurance-roadmap. risk assessment, model and system evaluation, red-teaming, external audit, documentation, data governance, monitoring, incident response and shared vulnerability reporting.
Support organisations’ access to and deployment of effective AI-enabled technologies to strengthen national resilience, workforce productivity and responses to emerging advanced threats.
As threats grow in speed, scale and sophistication, Australia's policy settings should support organisations' ability to access and deploy effective AI-enabled technologies, strengthening national resilience, workforce productivity and responses to emerging advanced threats.
Prioritise national-security and cyber-security AI investment through existing vehicles such as the National Reconstruction Fund to de-risk sovereign R&D, accelerate commercialisation and strengthen exports.
Prioritise public investment in AI in the national security and cyber security sectors through existing investment vehicles such as the National Reconstruction Fund to de-risk sovereign R&D, accelerate commercialisation and strengthen export channels.
Introduce incentives for cyber security uplift, expenditure and training and invest in the Infosec Registered Assessors Program (IRAP) for cyber readiness.
Introduce incentives for cyber security uplift, expenditure and training and invest in the Infosec Registered Assessors Program (IRAP) to ensure cyber readiness for an evolving AI-enabled threat landscape.
Co-design sector-specific AI cyber-threat guidance with industry and peak bodies, including agent authentication and accountable provenance when agents access systems, invoke tools or communicate across organisational boundaries.
Provide sector-specific guidance on AI-enabled cyber threats, co-designed with industry and peak bodies, including guidance on authenticating autonomous agents and establishing accountable provenance when agents access systems, invoke tools or communicate across organisational boundaries.