Submission 315
Chartered Accountants Australia and New Zealand (CA ANZ)
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.
Supports responsible AI adoption underpinned by public trust, professional judgement and accountability, from the perspective of the accounting profession. Recommends a nationally consistent, risk-based framework focused on high-risk uses, practical guidance for SMEs on data and responsibilities, AI literacy investment, and coordinated action on AI-enabled harms.
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.
Recommends a nationally consistent, risk-based and proportionate AI framework focused on high-risk use cases and unacceptable harms.
a nationally consistent, risk-based and proportionate AI framework focused on high-risk use cases and unacceptable harms.
Argues that regulation should target specific high-risk uses rather than treating all general-purpose AI systems or AI-enabled tools as high risk.
regulation should be practical, scalable and focused on specific high-risk use cases rather than treating all general-purpose AI systems or AI-enabled tools as high risk.
Recommends that Government clarify the responsibilities of developers, deployers and end users of AI, particularly where businesses rely on third-party software.
Government should clarify the responsibilities of developers, deployers and end users of AI, particularly where businesses rely on third-party software, and support adoption through plain-language guidance, education, templates and sector-specific risk assessment tools.
States that accountants are often among the most accessible professional advisers for regional small businesses and community organisations.
In regional communities, accountants are often one of the most accessible professional advisers for small businesses and community organisations.
Argues that any independent AI regulator should be coordinated with existing regulators to avoid overlap and duplicated compliance.
If an independent AI regulator is established, its role should be clearly coordinated with existing regulators to avoid regulatory overlap, inconsistent guidance and duplicated compliance obligations.
States that professionals will need stronger capabilities in AI literacy, data quality, sceptical review and human oversight.
professionals will require stronger capabilities in AI literacy, data quality, sceptical review, ethical reasoning and human oversight.
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 nationally consistent, risk-based and proportionate AI framework focused on high-risk use cases and unacceptable harms.
a nationally consistent, risk-based and proportionate AI framework focused on high-risk use cases and unacceptable harms.
Make AI guardrails practical and scalable, with clear responsibilities.
AI Guardrails that are practical and scalable with clear responsibilities.
Have AI policy support trusted advisers, confidentiality, cyber security, assurance, effective review and redress, and investment in AI literacy and digital capability.
AI policy should support trusted advisors, confidentiality, cyber security, assurance, effective review and redress, investment in AI literacy and digital capability
Continue investment in AI capability, governance and professional judgement to support productivity and living standards.
We recommend continued investment in AI capability, governance and professional judgement to support the uplift in productivity and our living standards.
Recognise the trusted adviser role in government AI policy as part of the delivery infrastructure needed to extend AI capability beyond metropolitan centres.
This trusted adviser role should be recognised in government AI policy as part of the delivery infrastructure needed to extend AI capability beyond metropolitan centres.
Support responsible adoption through plain-language AI playbooks, governance toolkits and practical training for SMEs, regional businesses and the professions who advise them.
Government should support responsible adoption through plain-language AI playbooks, governance toolkits and practical training for SMEs, regional businesses and the professions who advise them.
Clarify the responsibilities of developers, deployers and end users of AI, particularly where businesses rely on third-party software, and support adoption through plain-language guidance, education, templates and sector-specific risk assessment tools.
Government should clarify the responsibilities of developers, deployers and end users of AI, particularly where businesses rely on third-party software, and support adoption through plain-language guidance, education, templates and sector-specific risk assessment tools.
Develop practical, scalable AI data guidance for all sectors, with specific support for SMEs, covering where data is stored, whether it trains AI models, who can access it, and applicable obligations.
We recommend that government develop practical, scalable AI data guidance for all sectors, with specific support for SMEs. Guidance should help organisations understand where data is stored, whether it is used to train AI models, who can access it, and what privacy, confidentiality, cyber security and professional conduct obligations apply.
Invest ongoing in education, professional development and lifelong learning for all professions so AI augments professional expertise.
Ongoing investment in education, professional development and lifelong learning for all professions will be essential to ensure AI augments professional expertise, improves productivity and supports the workforce needed by Australia’s economy.
Make any AI legislative framework targeted, proportionate and coordinated with existing laws, clarifying obligations of developers, deployers, end users and offshore providers while avoiding unnecessary duplication.
any legislative framework should be targeted, proportionate and coordinated with existing laws. It should clarify obligations for developers, deployers, end users and offshore providers, while avoiding unnecessary duplication with existing privacy, consumer protection, corporations, tax, audit, insolvency, cyber security, anti-money laundering and professional standards frameworks.
Limit any approval process for high-risk AI use cases to applications with significant legal, financial, employment, human rights or safety impacts, not low-risk productivity tools or unnecessary SME delay and cost.
Any approval process for high-risk AI use cases should be limited to applications with significant legal, financial, employment, human rights or safety impacts. It should not capture low-risk business productivity tools or impose unnecessary delay and cost on SMEs and accounting practices.
If an independent AI regulator is established, coordinate its role with existing regulators, with practical guidance and enforcement focused on high-risk and harmful AI applications.
If an independent AI regulator is established, its role should be clearly coordinated with existing regulators to avoid regulatory overlap, inconsistent guidance and duplicated compliance obligations. The regulator should provide practical guidance, support regulatory coherence and focus enforcement on high-risk and harmful AI applications.
Clarify expectations for AI software so the public can undertake proportionate due diligence, maintain oversight and understand responsibility for privacy, confidentiality, IP, cyber security and AI-assisted outputs.
We recommend that government clarify expectations for AI software used so the public can undertake proportionate due diligence, maintain oversight and understand responsibility for privacy, confidentiality, intellectual property, cyber security and AI-assisted outputs.
Coordinate government, regulators, technology providers, financial institutions and professional bodies to reduce AI harms through education, guidance, verification, stronger reporting, safeguards and effective complaints, review and redress.
We recommend coordinated action across government, regulators, technology providers, financial institutions and professional bodies to reduce AI-enabled harms. This should include public education, business and regulatory guidance, verification tools, stronger reporting pathways, safeguards, and effective complaints, review and redress mechanisms.
Ensure regulators, the Australian AI Safety Institute and the intelligence and security community have appropriate technical expertise, resourcing and information-sharing arrangements.
Regulators, the Australian AI Safety Institute and the intelligence and security community should have appropriate technical expertise, resourcing and information-sharing arrangements.
Primarily assign review and testing of powerful general-purpose models and advanced systems to developers and providers; inform downstream users sufficiently for proportionate risk assessment, professional judgement, controls and assurance.
We recommend that review and testing obligations for powerful general-purpose AI models and advanced systems primarily apply to developers and providers best placed to assess capability, limitations, safety and security. Downstream users, including accounting firms and finance teams, should receive sufficient information to undertake proportionate risk assessments and apply appropriate professional judgement, controls and assurance.