This article is part 4 of our series about ‘Law in the AI age’, where we provide our position on how Australia’s legislative framework should be reformed to address the opportunities and challenges presented by artificial intelligence. It is our response to the Government’s Productivity Commission Interim Report on ‘Harnessing Data and Digital Technology’, published in August 2025.  

Our series explores emerging legal issues with respect to genAI, including legal liability for AI advice, whether users of genAI platforms need to disclose their chats to a Court, and whether the Copyright Act should be amended to allow the use of copyright works for training purposes.  

Background

The Productivity Commission (PC) is the Australian Government’s independent research and advisory body. In August 2025, the PC released the Harnessing Data and Digital Technology Interim Report (Interim Report). The Interim Report explores the opportunities and challenges presented by AI in Australia, including its potential to boost the productivity of Australian workers, and the regulatory issues it raises.

This article outlines Edwards + Co Legal’s formal position on how Australia’s legislative framework should be reformed in response to the PC’s Interim Report. Edwards + Co Legal is a boutique corporate and commercial law firm based in Sydney, specialising in media, digital technologies, intellectual property and new-economy businesses. We regularly advise clients at the intersection of law, technology and innovation. We are legal experts on how regulation can drive innovation, protect creators, and strengthen Australia’s global competitiveness.

We submit that a balanced, commercially informed, and legally sound approach is required, that supports both productivity gains whilst also sustaining the rights of creators.

Opportunities for Business and the Economy

For modern Australian business, AI is no longer a future frontier, it is here now.  AI stands to significantly enhance productivity across Australian industries. Businesses report AI’s immediate value in streamlining administrative work, improving drafting and summarisation workflows, and unlocking advanced data analysis.  According to a Business Council of Australia report released in June 2025, 90% of SMEs using AI report more efficient operations.[1] The report also states that generative AI for high-skilled work reduces perceived mental demand by 45%, and perceived stress and difficulty by 58%. The PC’s estimate of up to AU$116 billion in GDP uplift over the next decade underscores this potential.

To encourage Australian businesses to scale AI, legal clarity is required to avoid the perception of risks, which would have a limiting effect on economic benefits.

ISO/IEC 42001:2023 is the current standard for artificial intelligence management system (AIMS). It provides a structured framework for AI governance, which helps organisations build trust, achieve AI compliance and align with international best practices. The goals of this standard are responsible development, deployment and operation; critical factors in successful AI adoption and broader transformation.

Balancing Innovation with the Rights of Creators

Current Australian Copyright Law

Under Australian copyright law, copyright arises in original literary, dramatic, musical and artistic works that have been reduced to material form, provided that there is sufficient independent intellectual effort by a human creator. The copyright owner has the exclusive rights set out in section 31 of the Copyright Act 1968 (Cth) (Copyright Act), to use the copyright material in certain ways and control the use of it by anyone else.  A copyright owner has the right to licence others to use their copyright, or to transfer/assign it.

Copyright infringement occurs when work is used in one of the exclusive ways controlled by the copyright owner, without the copyright owner’s permission.[2] This includes where a ‘substantial part’ of the material is used. However, there are certain exceptions under the Copyright Act, where copying is permitted.

Compiling data sets to train AI requires the AI model to at least temporarily reproduce the works, and it is technically an infringement of copyright to reproduce protected works without authorisation from the copyright holder. LLMs need to come within one of the other exceptions under Australian law. The current fair dealing exceptions were not designed to apply to the training of AI models, and their application to such uses remains legally uncertain.

We recognise the legitimate rights of creators, such as authors, artists, and rights holders, to control the use of their works in AI training. Without fair compensation for innovation, creators lack incentives to produce the new material that the creative ecosystem requires to thrive. This is even more so the case in a world where AI models require data inputs in the terabytes.

Copyright Act Reform

To address this, we recommend the amendment of the Copyright Act to include a new fair dealing exception that allows the use of copyright material for targeted text and data mining (TDM) for AI training purposes.

However, we consider that the TDM exception should be subject to some key conditions:

• Transparency in data sources: LLM owners/controllers using copyrighted material for AI training or other TDM purposes must clearly disclose what data they used and where it came from. This may be achieved via public disclosure of the datasets used in the training.
• Opt-out mechanisms or licensing paths for rights holders: Rights holders could formally tag their works as not available for AI training, such as through a metadata tag or registration with a centrally managed opt-out list. Alternatively, some owners or aggregation agencies, could negotiate licences, enabling lawful use of their works in return for consideration.
• Proportional remuneration frameworks for rights holders: A balanced approach with appropriate compensation for use of copyright works that reflects the scale and commercial value of use. A small AI startup scraping a few articles is not to be treated the same as a major AI player training on millions of IP works. Collective licensing schemes (like APRA/AMCOS for music rights) could be considered, where creators receive distributions based on AI usage share.

This approach preserves legal certainty for businesses and safeguards creators’ commercial interests, fostering shared growth.

Regulatory Principles and Legal Foundations

We advocate the following legal principles:

1.  Light-Touch, Risk-Based Reform

• It is not necessary for entirely ‘greenfield’, AI-specific legislation to be introduced. Instead, we should work with the existing statutory and regulatory frameworks – in particular the Privacy Act, Copyright Act, and Competition and Consumer Act.
• Australia already has strong laws, AI is evolving too fast for rigid regulation, whilst international best practice favours adapting existing frameworks. As a matter of lawmaking policy, new legislation should be a last resort if proven gaps emerge.

2.  Technology-Neutral by Legal Design

• To the extent that we can, reforms and amendments should be drafted in a technology-agnostic, future-proof way. Reform should not be tied only to AI technology developments.
• This ensures that legal frameworks regulate outcomes and risks rather than particular technologies. Reforms should not be tied exclusively to AI developments, but should apply consistently across evolving digital technologies that raise similar regulatory issues. Such an approach provides durability, reduces compliance costs, and aligns with global best practice.

3.  Alignment with International Standards

• Encourage adoption of recognised AI governance standards (e.g. ISO 42001) to support cross-border operability and export competitiveness.
• Such standards support interoperability with global markets, enhance the export competitiveness of Australian businesses, and provide a practical pathway for demonstrating responsible AI practices without imposing duplicative regulation.

4.  Outcomes-Focused Privacy Regulation

• Shift from prescriptive, process-driven requirements to outcome-based regulation.
• Make businesses accountable for achieving regulatory goals, such as transparency, safety and fairness, without being overburdened by outdated or overly rigid processes.
• An outcome-based approach provides flexibility for innovators, reduces compliance costs, and aligns with modern regulatory expectations internationally.

Recommended Reforms

In summary, Edwards + Co Legal recommends the following reforms in response to the PC’s Interm Report:

1. Introduce a narrowly tailored TDM exception in Copyright Act, with fairness safeguards: opt-out provisions, transparency requirements, and licensing options.
2. Encourage voluntary AI governance through frameworks such as ISO 42001, including through guidance or incentives.
3. Pilot government use cases, where the public sector leads in responsible AI deployment, providing precedents for private-sector adoption.
4. Support programs to promote AI literacy and capability-building, via industry collaboration or funding to address workforce gaps in AI knowledge and skills, better equipping them for the modern LLM era.
5. Establish a streamlined AI oversight mechanism, acting as inter-agency coordinator and repository for guidance, to reduce regulatory fragmentation.

Legal Perspective and Precedents

Our firm’s work advising on digital, media, and IP matters has demonstrated that legal certainty drives commercial confidence, particularly in emerging technology. For both creators and digital clients alike, uncertainty around the legality of AI training significantly dampens investment.

Moreover, recent digital reforms reflect how outdated legal assumptions were successfully modernised without sacrificing safeguards. For example, with the momentum brought about by the COVID-19 pandemic, the Corporations Act was amended in 2022 to permanently authorise electronic execution, virtual shareholder meetings and electronic issue of documents to shareholders. Other modernising developments include the ability to lodge forms electronically with ASIC.

AI regulation should be similarly adaptive.

Concluding Remarks

In Harnessing Data and Digital Technology, the PC has taken a forward-looking approach. Edwards + Co Legal welcomes an outlook that seeks to seize AI’s productivity potential, whilst balancing creators rights to IP. This is crucial to Australia’s future business competitiveness and cultural relevance.

Our proposed reforms of the Copyright Act advances pragmatic reforms that are both business-focused and IP-conscious. Our recommendations are designed to reduce regulatory complexity, provide clarity for original content creators, and establish frameworks that facilitate responsible growth while ensuring innovators who create or own original works receive due consideration. We emphasise reforms that are practical to implement, internationally aligned, and capable of adapting to emerging technologies.

Does your business use artificial intelligence? For specific advice on how to use and commercialise AI in your business, please contact us below. Edwards + Co Legal provide corporate and commercial legal advice to modern Australian businesses.

[1] Business Council of Australia, Accelerating Australia’s AI Agenda (https://aiagenda.bca.com.au/wp-content/uploads/2025/05/238_AI-Report_FINAL_WEB.pdf?)

[2] Copyright Act 1968 (Cth), s 36.