Submission 42
Dr Daniela Simone, Professor Isabella Alexander, Professor Kathy Bowrey, Professor Michael Hander, Dr Sarah Hook, and Professor Kimberlee Weatherall
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.
Six legal academics argue that the Copyright Act 1968 cannot fully deal with AI. They describe infringement risks beyond training, limits of licensing, creators' barriers to justice, the need for research exceptions, and the need for transparency, and recommend law reform plus remedies and remuneration alongside copyright.
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 infringements could arise at multiple points in AI development and operation, with a range of parties potentially jointly or severally liable.
Infringements could arise at multiple points during development and operation of AI systems. A range of parties may be jointly, or severally, liable for those infringements.
Argues that moral rights are owned by individuals, cannot be negotiated in industry AI licensing agreements and cannot be waived through broad contractual provisions.
Moral rights are owned by individuals and cannot be negotiated in industry-led AI licensing agreements. Importantly, they cannot be waived through broad contractual provisions, as recently confirmed by the Federal Court.
Argues that no clear legal requirement ensures authors receive economic benefits, or even notice, when AI companies and rightsholders negotiate deals including their works.
There is no clear legal requirement for ANY economic benefits to be provided to authors when negotiations occur between AI companies and rightsholders. It is not clear if authors even need to be advised when their works are included in AI deals.
Says enforcement routes accessible to individual creators, such as a small claims forum, seem promising and deserve full consideration.
Routes for enforcement that are accessible to individual creators (who are often small business owners), such as the proposal to create a “small claims” forum, as exists in other jurisdictions seem promising and deserve full consideration, perhaps alongside better complaints and takedown services by platforms.
Recommends reforming copyright exceptions to enable Australian research into AI, using AI in other fields, and understanding how AI works.
copyright exceptions need to be reformed to enable cutting edge Australian research (i) into AI itself; (ii) using AI to make advances in a variety of other fields; and (iii) directed to understanding how AI works.
Argues that transparency and accountability questions go beyond copyright, and that the transparency obligations needed for AI go well beyond those needed to administer a licensing regime.
But questions around transparency and accountability clearly transcend the copyright context. The transparency obligations necessary for AI go well beyond those necessary to administer a licensing regime.
Recommendations
Recommendations are shown beside the submission’s own words. Listed in the same order as on its page in the interactive guide.
Provide additional remedies and remuneration for Australian creators alongside copyright.
We recommend additional remedies and remuneration for Australian creators alongside copyright.
Review Copyright Act definitions and provisions concerning copying, communication, substantial parts, reproduction, articles, computer-program exceptions, research or study, and moral rights in relation to AI operations and their interaction.
Current provisions and definitions in the Copyright Act 1968 (Cth) that should be reviewed with respect to AI operations and interaction with each other include: the s 10(1) definitions of “copy”, “infringing copy”, “computer program” and “communicate”; s 14(1) on “substantial part”; s 21 on “reproduction”; ss 37(1) and 38(3) on “article”; Part 3, Division 4A on acts not constituting infringements of copyright in computer programs; s 40 on fair dealing for the purpose of research or study; and Part IX Divisions 1, 2, 2A, 3, 3A, 4, 4A, 6 on moral rights.
Require permission and fair remuneration for use of Indigenous Cultural and Intellectual Property, taking account of First Nations bargaining vulnerabilities and copyright inadequacies.
ICIP should not be used without permission and fair remuneration. The vulnerabilities of First Nations people in bargaining situations and the inadequacy of existing copyright in protecting ICIP need to be kept front of mind.
Assess creators’ practical ability to use any proposed Copyright Act solutions.
Any changes to the Copyright Act 1968 (Cth) requires a consideration of the practicalities of creators being able to use any proposed solutions.
Give full consideration to enforcement routes accessible to individual creators, such as a small claims forum, perhaps alongside better platform complaints and takedown services, and to creator protections beyond copyright.
Routes for enforcement that are accessible to individual creators (who are often small business owners), such as the proposal to create a “small claims” forum, as exists in other jurisdictions seem promising and deserve full consideration, perhaps alongside better complaints and takedown services by platforms. Solutions to protect creator interests that go beyond copyright and its enforcement are also needed.
Update Australia's copyright exceptions for the digital environment and AI, including broader exceptions for education, publicly funded research and other public interest uses.
If the Australian government fails to amend the law to provide broader exceptions for education, publicly funded research and other public interest uses of copyright content – exceptions which exist overseas
Reform copyright exceptions to enable research into AI itself, AI-assisted advances in other fields, and understanding how AI works.
copyright exceptions need to be reformed to enable cutting edge Australian research (i) into AI itself; (ii) using AI to make advances in a variety of other fields; and (iii) directed to understanding how AI works.
Amend the fair dealing exception for research so it covers research into AI systems, their operation and interoperability.
Amending the existing limited fair dealing exception for research to ensure that it at least encompasses a right to research AI systems, operations and interoperability is an important step in improving transparency, by allowing Australian researchers to understand what AI models are doing and to help us build our own.
Clarify the scope of copyright exceptions for security testing, interoperability and incidental or technical use, in light of AI and data centre operations.
Finally, there is a need to clarify the scope of copyright exceptions for security testing, interoperability and incidental or technical use (designed for web hosting and caching), in light of AI and the diversity of data centre operations and AI applications.
Implement a right to research AI systems and their operation and impacts on society through focused exceptions reform.
Focused exceptions reform is needed to remove impediments to research AI technologies by implementing a right to research AI systems and their operation and impacts on society.
Prevent contracting out of copyright exceptions.
Contracting out of exceptions should be prevented to ensure their effectiveness.
If licensing is adopted, require transparency of collecting societies, deal structures, granular material-access logs, payment formulas, and training datasets and their use.
If a licensing solution is adopted, greater transparency is needed around the activities and operation of collecting societies. Information will be needed about the structure and operation of licensing deals. This includes logs of when, how and how often protected material is licensed or accessed by AI developers at a granular level of detail. The formulas (or proxies) used to calculate payments should also be disclosed. Whilst transparency around the content of training datasets and their use is necessary to underpin any licensing solution, this is just one part of a broader conversation about the information needed to regulate AI more generally.
Require AI transparency and accountability beyond the information needed to administer copyright licensing.
But questions around transparency and accountability clearly transcend the copyright context. The transparency obligations necessary for AI go well beyond those necessary to administer a licensing regime.
Assess any copyright reform for coherence with other areas such as privacy, e-safety, competition, anti-discrimination, employment law and sector-specific regulation.
Any reform contemplated should be assessed for its coherence with other areas (eg privacy, e-safety, competition, anti-discrimination, employment law, sector specific regulation).