Submission 307
Australian Banking Association (ABA)
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 AI is a generational productivity opportunity and that regulation should be technology neutral, proportionate and build on existing frameworks first, with new AI-specific obligations only where a gap is demonstrated. Asks that industry's use of AI against scams and fraud be preserved, and recommends support for skills, SME adoption and domestic 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.
Says independent modelling estimates AI could add $95–116 billion to Australian GDP by 2036 and lift multifactor productivity.
Independent modelling estimates that Artificial Intelligence (AI) could add $95 billion to $116 billion to Australia’s GDP by 2036, lift multifactor productivity by 2.0 to 2.4 per cent and create a net 36,000 to 44,000 full-time equivalent jobs.
Argues that banking already operates under extensive technology-neutral regulation covering many risks attributed to AI.
Banking already operates under an extensive body of technology neutral regulation covering the risks most often attributed to AI, such as, discrimination and bias, financial exclusion, privacy and data protection, transparency and explainability, cyber security, responsible lending, accountability, and third- party risk.
Argues that broad high-risk definitions could unintentionally capture fraud, scam and financial crime detection.
Broad high-risk definitions could unintentionally capture fraud, scam and financial crime detection, and the identification of financial abuse or customer vulnerability.
States that banks want to use Australian AI but that this depends on a strong domestic ecosystem of providers, researchers and skilled workers.
Banks want to use Australian technology, Australian expertise and Australian innovation, but this depends on a strong domestic ecosystem of AI providers, researchers and skilled workers
Proposes a verifiable digital identity to counter AI-fabricated documents in loan fraud.
A verifiable digital identity, available across the digital economy rather than only for government services, would help banks establish who an applicant is without relying on a document AI could fabricate.
Argues that transparency obligations on automated decisions should not extend to disclosing how fraud and financial crime detection systems work.
From December 2026, where an automated decision involving personal information significantly affects a customer’s rights or interests, banks will be required to disclose the use of automated decision-making in general terms.16 This should not extend to disclosing how fraud and financial crime detection systems work, which would allow them to be evaded.
Recommendations
Recommendations are shown beside the submission’s own words. Listed in the same order as on its page in the interactive guide.
Prioritise innovation, investment and AI adoption across the economy.
Prioritise innovation, investment and the adoption of AI across the economy to unlock productivity growth, improve customer outcomes and strengthen Australia's domestic capability.
Support Australians and businesses of all sizes through skills, training, reskilling, targeted SME assistance and responsible AI use to improve accessibility and inclusion.
Support Australians and businesses of all sizes to participate in the AI transition through skills, training and reskilling initiatives, targeted assistance for small and medium businesses, and the responsible use of AI to improve accessibility and inclusion.
Expressly preserve industry's ability to use AI against scams, fraud and cyber-crime, reinforced by reforms to digital identity and Government data sharing.
Expressly preserve industry's ability to use AI to protect Australians and Australian businesses from scams, fraud and cyber-crime. This should be reinforced by reforms to digital identity and to the sharing of Government data.
Regulate technology neutrally by risk and outcome rather than technical characteristics, applying the same requirements to the same risks across the economy.
Be technology neutral, regulating risks and outcomes rather than the technology itself, and applying the same requirements to the same risks across the economy.
Before new obligations on already regulated sectors, identify and publish gaps between existing law and the risk; legislate only where a gap is established.
The ABA asks the Committee to recommend that, before new obligations are imposed on an already regulated sector, the Government identify and publish the gaps between existing law and the risk it is seeking to address and legislate only where a gap is established.
Introduce new AI-specific obligations only after identifying the risk, assessing existing legal, regulatory and prudential coverage, demonstrating inadequacy and a clear gap, and ensuring proportionality to the risk’s nature and scale.
The ABA's position is that new AI-specific obligations should be introduced only after the following steps are undertaken: 1. Identify the risk it is seeking to address. 2. Assess whether existing legal, regulatory and prudential frameworks already address that risk. 3. Demonstrate a clear regulatory gap where existing frameworks are inadequate. 4. Ensure any new obligations are proportionate to the nature and scale of the risk.
Ensure any additional regulation is proportionate to risk and customer impact, reflects the existing regulatory environment, and does not require disclosure that would undermine fraud or financial-crime detection systems.
Ensure that any additional regulation is proportionate to the risk and its impact on customers, reflects the existing regulatory environment, and does not require disclosure that would undermine fraud or financial-crime detection systems.
Build the domestic AI ecosystem through investment in compute infrastructure, research, commercialisation and AI skills, using procurement and other policy levers to help Australian firms scale.
Government can help build that ecosystem through investment in compute infrastructure, research, commercialisation and AI skills, and by using procurement and other policy levers to support Australian firms to scale.
Ensure regulatory settings enable banks to invest in domestic compute, research and AI skills, and use Commonwealth procurement to build a market for Australian AI providers.
ensure that regulatory settings enable banks to invest in domestic compute, research and AI skills, and use Commonwealth procurement to build a market for Australian AI providers.
Build AI skills and reskilling into Australia's existing vocational and tertiary training pathways.
The Committee should support AI skills and reskilling being built into Australia's existing vocational and tertiary training pathways.
Recognise the gains AI offers to customer outcomes, including accessibility and inclusion, where appropriate safeguards and access to human support are maintained.
The Committee should also recognise the gains AI offers to customer outcomes, including accessibility and inclusion for customers, where appropriate safeguards and access to human support are maintained.
Support a verifiable digital identity available across the digital economy, and sharing of ATO income data through the Consumer Data Right, to counter AI-fabricated loan documents.
A verifiable digital identity, available across the digital economy rather than only for government services, would help banks establish who an applicant is without relying on a document AI could fabricate. Sharing Australian Taxation Office income data through the Consumer Data Right would do the same for what an applicant earns, cross-checking income at its source. Together they would close off the two inputs where AI could have a negative impact, and the ABA asks the Committee to support both.
Preserve AI detection and prevention of scams, fraud, financial crime, financial abuse and customer vulnerability, excluding these uses from high-risk designations or AI-specific obligations.
The ABA asks the Committee to recommend that the Government expressly preserve industry's ability to use AI to detect and prevent scams, fraud, financial crime, financial abuse and customer vulnerability, so that no high-risk designation or AI-specific obligation captures those uses.
Develop any new AI rules with regard to international interoperability.
Where new rules are needed, they should be developed with regard to international interoperability.
Have future and in-flight reforms consider their applicability to AI to avoid regulatory gaps emerging.
Future and in-flight reforms should consider their applicability to AI to avoid regulatory gaps emerging in the future.
Ensure the SPF, PSP licensing and ePayments Code can manage agentic commerce’s potentially greater scam, fraud and consumer-protection risks.
Agentic commerce that enables AI agents to transact on a customer’s behalf is expected to amplify existing payment system risks. The SPF, PSP licensing and ePayments Code will need to be fit for purpose to manage potentially greater scams, fraud and consumer protection risks.
Consider competition, resilience and sovereignty risks from AI platforms’ growing economic role.
Risks to competition, resilience and sovereignty will also need to be considered from the growing role of AI platforms in the economy.
Reduce unnecessary data collections and duplicative reporting.
The same principle underlies the ABA's advocacy on regulatory efficiency, including its recommendation to reduce unnecessary data collections and duplicative reporting.
Do not require businesses to expose proprietary algorithms or intellectual property through accountability or transparency policies.
The ABA has said that "accountability or transparency policies should not require businesses to expose proprietary algorithms or IP"
Set transparency and oversight obligations in proportion to customer impact.
The ABA asks the Committee to recommend that transparency and oversight obligations be set in proportion to the impact on the customer.