Submission 447
Mr David Hynes
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.
An individual who runs SME AI adoption workshops proposes a national AI media, literacy and dialogue institution as a flagship recommendation, with packages on creator and content integrity, community safety, adoption and sovereign capability, and governance. Measures include training-data transparency, deepfake takedown, SME advisory clinics delivered through the institution, and a register of Commonwealth AI use.
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.
Says National AI Centre survey data shows SME non-adopters’ distrust and uncertainty, and argues distrust is an information and institutions problem.
The National AI Centre's survey (December 2025–February 2026) shows the consequence: SME adoption sits at 43–44%, and among non-adopters ~65% cite distrust or a preference for human control, 54% cannot see AI's relevance, and 19% do not know where to start [1]. Distrust is not a technology problem; it is an information and institutions problem.
Argues that no Australian institution is resourced to explain AI in plain language, report on it independently and carry Australian voices back to decision-makers.
No Australian institution is resourced to explain AI in plain language, report on it independently, and carry Australian voices back to decision-makers.
Recommends regulator-mediated training-data transparency for commercial models and support for collective licensing marketplaces.
The Committee should recommend: (a) regulator-mediated training-data transparency for commercial models offered in Australia, with rightsholder query rights and no forced trade-secret disclosure; (b) support for collective licensing marketplaces so individual and small rightsholders can license practically;
Recommends extending scams obligations to AI-enabled impersonation and a funded eSafety takedown pathway for deepfake abuse and AI scam content with a 48-hour standard.
The Committee should recommend: (a) extending Scams Prevention Framework obligations on banks, telcos and platforms to AI-enabled impersonation (voice clones, synthetic executives and relatives), with AI-detection measures in high-risk channels; (b) a funded eSafety takedown pathway for deepfake abuse and AI scam content with a 48-hour standard;
Says it does not ask the Committee to revisit the decision against economy-wide guardrails, proposing instead the Commonwealth's own AI use and purchasing power and targeted gap-closing in defined high-risk settings.
This submission does not ask the Committee to relitigate that decision economy-wide. Two enforceable anchors are available now: the Commonwealth's own AI use and purchasing power (TOR (h)), and targeted gap-closing in defined high-risk settings (TOR (i))
Recommendations
Recommendations are shown beside the submission’s own words. Listed in the same order as on its page in the interactive guide.
Establish a national AI media, literacy and dialogue Hub as an arm’s-length public-interest institution.
establish a national AI media, literacy and dialogue institution ("the Hub") — an arm's-length public-interest body
Provide independent AI news, explainers, incident reports and verification guidance for elections, emergencies and scams, with free open republication by Australian publishers and ABC/SBS partnerships guaranteeing regional reach.
Independent Australian AI news, explainers, incident reporting and verification guidance for elections, emergencies and scam waves — produced for free syndication to Australian publishers under an open-republication model (as pioneered by The Conversation), supporting media sustainability (TOR (j)); ABC/SBS distribution partnerships guarantee regional reach (TOR (b)).
Create a national AI literacy program adapted from the UNESCO AI competency frameworks, with curriculum-mapped school resources and funded teacher professional development.
A national program adapted from the UNESCO AI competency frameworks for students and teachers [5]: curriculum-mapped school resources with funded teacher professional development
Offer adult AI literacy through TAFE micro-credentials, libraries and employer co-funding, plus seniors’ scam and deepfake resilience programs with eSafety.
adult pathways through TAFE micro-credentials, libraries and employer co-funding, echoing the EU AI Act's Article 4 literacy duty [6]; and seniors' scam- and deepfake-resilience programs with eSafety.
Provide that student data is not training data and that humans decide high-stakes outcomes.
Two protections should follow enacted US state law (Idaho SB 1227; California AB 1159): student data is not training data, and humans decide high-stakes outcomes [7].
Hold quarterly town halls in every state and territory, prioritising regional, remote, outer-suburban, First Nations and multicultural communities, plus citizens' panels chosen by lot.
Quarterly town halls in every state and territory — prioritising regional, remote, outer-suburban, First Nations and multicultural communities — plus citizens' panels chosen by lot
Table an annual Australian AI Conversation report in Parliament and feed its findings to the AI Safety Institute, regulators and the R5 register as a standing public-evidence channel.
The annual "Australian AI Conversation" report is tabled in Parliament and feeds the AI Safety Institute, regulators and the R5 register — a standing public-evidence channel for future inquiries instead of a four-week submissions window (TOR (h)).
Give the Hub an independent board of educators, creators, small business, First Nations representatives, safety and technical experts, an editorial charter and no funder editorial control.
Arm's-length governance: an independent board (educators, creators, small business, First Nations representatives, safety and technical experts), an editorial charter, and no funder holding editorial control.
Have the Hub commission and coordinate with the National AI Centre, eSafety, ABC/SBS, TAFE, libraries, chambers of commerce and regional organisations.
A hub, not an empire: it commissions and coordinates with the National AI Centre, eSafety, ABC/SBS, TAFE, libraries, chambers of commerce and regional organisations.
Provide Commonwealth base funding and diversified supplements without influence; measure literacy, reach, syndication, town-hall participation and movement in National AI Centre adoption data.
Commonwealth base funding with diversified supplements that buy no influence. Measured by literacy surveys, reach and syndication uptake, town-hall participation, and movement in the National AI Centre's adoption data [1].
Maintain licensing-first copyright.
R2 — Creator and content-integrity package: maintain licensing-first copyright;
Establish regulator-mediated training-data transparency for commercial models offered in Australia, with rightsholder query rights and no forced trade-secret disclosure.
regulator-mediated training-data transparency for commercial models offered in Australia, with rightsholder query rights and no forced trade-secret disclosure;
Support collective licensing marketplaces so individual and small rightsholders can license practically.
support for collective licensing marketplaces so individual and small rightsholders can license practically;
Introduce a digital replica/personality right covering face, voice and style with civil and criminal remedies, complementing existing deepfake offences and including ICIP protocols requiring free, prior and informed First Nations consent.
an Australian digital replica / personality right (face, voice, style) with civil and criminal remedies, complementing the 2024 deepfake sexual material offences [13] and including ICIP protocols requiring free, prior and informed consent for First Nations content;
Introduce national AI content labelling and provenance standards (EU AI Act Article 50; C2PA-style credentials).
a national AI content labelling and provenance standard (EU AI Act Article 50 [6]; C2PA-style credentials)
Establish a verified official communications channel for banks, government and emergency alerts.
standard (EU AI Act Article 50 [6]; C2PA-style credentials) plus a verified official communications channel for banks, government and emergency alerts;
Establish a standing AI-Creative Industries Dialogue and grants helping Australian creators and media businesses use AI.
a standing AI–Creative Industries Dialogue and grants helping Australian creators and media businesses use AI, so the sector features in the opportunity story too.
Extend Scams Prevention Framework obligations on banks, telcos and platforms to AI-enabled impersonation, with AI-detection measures in high-risk channels.
extending Scams Prevention Framework obligations on banks, telcos and platforms to AI-enabled impersonation (voice clones, synthetic executives and relatives), with AI-detection measures in high-risk channels;
Fund an eSafety takedown pathway for deepfake abuse and AI scam content with a 48-hour standard.
a funded eSafety takedown pathway for deepfake abuse and AI scam content with a 48-hour standard;
Subsidise AI-security tooling for SMEs.
subsidised AI-security tooling for SMEs — 48% of SME AI adopters already use AI for cybersecurity [1], so policy should scale what works;
Impose age-appropriate duties on AI companions and chatbots likely accessed by minors: age assurance, crisis responses to self-harm or abuse disclosures, and no romantic or sexualised engagement with minors.
age-appropriate design duties for AI companion and chatbot apps likely accessed by minors: age assurance, crisis-response protocols for self-harm or abuse disclosures, and no romantic or sexualised engagement with minors.
Have school curricula cover AI-mediated harassment and image-based abuse.
school curricula should cover AI-mediated harassment and image-based abuse.
Establish SME Trusted Advisor clinics delivered through the Hub's local network via chambers of commerce, accounting bodies, industry associations and regional organisations.
SME Trusted Advisor clinics delivered through the Hub's local network via chambers of commerce, accounting bodies, industry associations and regional organisations,
Provide sector playbooks for lagging industries.
with sector playbooks for lagging industries (construction and agriculture sit below 30% adoption [1])
Provide a plain-English 'first 90 days' safe-use guide aligned to the Guidance for AI Adoption.
a plain-English "first 90 days" safe-use guide aligned to the Guidance for AI Adoption
Create a sovereign compute access program giving SMEs, startups and researchers credits toward domestic or Australian-data-resident compute.
a sovereign compute access program giving SMEs, startups and researchers credits toward domestic or Australian-data-resident compute;
Create an Australian-Data Models Program through data trusts and cooperatives for agriculture, resources, health and medical research, advanced manufacturing, financial services and defence-adjacent dual-use sectors.
an Australian-Data Models Program for sectors where Australian data is a competitive asset —agriculture, resources, health and medical research, advanced manufacturing, financial services, defence-adjacent dual-use — delivered through data trusts and cooperatives [3];
Use Commonwealth procurement as anchor customer (pilot quotas, challenge-based buying).
Commonwealth procurement as anchor customer (pilot quotas, challenge-based buying), converting TOR (h) into industrial policy;
Introduce talent levers: national AI fellowships, R&D tax incentives extended to compute, and commercialisation bridges from university labs.
talent levers: national AI fellowships, R&D tax incentives extended to compute, and commercialisation bridges from university labs.
Make conformance with the Guidance for AI Adoption and Australia's AI Ethics Principles a mandatory condition of Commonwealth AI procurement and use.
conformance with the Guidance for AI Adoption and Australia's AI Ethics Principles as a mandatory condition of Commonwealth AI procurement and use,
Maintain a public register of Commonwealth AI use-cases (risk classification, human oversight).
a public register of Commonwealth AI use-cases (risk classification, human oversight);
Table periodic AI Safety Institute reports in Parliament on capabilities, incidents and emerging risks, with the Hub's town-hall findings as the community-input stream.
periodic AI Safety Institute reports tabled in Parliament on capabilities, incidents and emerging risks, with the Hub's town-hall findings as the community-input companion stream;
Conduct an 18-month review of regulatory gaps in hiring, credit, health diagnostics, policing and children’s services.
an 18-month targeted review of regulatory gaps in defined high-risk settings (hiring, credit, health diagnostics, policing, children's services);
Embed customer-facing transparency practices (disclose AI interaction; a route to raise concerns) in SME guidance.
customer-facing transparency practices (disclose AI interaction; a route to raise concerns) embedded in SME guidance under R4