Response to Safe and Responsible AI In Australia Discussion Paper

1.    The Queensland Council for Civil Liberties (“the QCCL”) is a not-for-profit organisation that promotes civil liberties and receives queries from members of the public regarding their civil liberties and individual rights.

2.    We make this submission in response to the Safe and responsible AI in Australia Discussion Paper (“the Discussion Paper”).

3.    Our submission is straightforward - we consider that discussion regarding artificial intelligence (in any form) should be focused upon the best interest of individual human rights. Australia is unique in that it has refrained from legislating human rights protections and our Federal Constitution explicitly lacks human rights[1]. In this context, we consider that a human rights framework[2] ought to be developed prior to (or at least alongside) any development of a framework for governing the use of advanced computational processes as these developments have the potential to seriously impact individual freedoms. 

4.    In our view, Australia should require any development or application of artificial intelligence to an authorisation and licensing process that primarily focuses on the human rights implication(s) and we submit that this approach ought to be consistently applied to State use of these computational processes as well as any private development or application.

5.    We consider that the approach being adopted in the EU AI Act[3] offers a sensible guidance to this process and that, in that context, Australia ought to have regard not only to the terms of that legislative development but also to the significant human-centric and human rights focused norms that underpin European jurisprudence.

6.    In particular, we consider that the General Obligations contained in Chapter II of the EU AI Act provide a reasonable and acceptable primary purpose for any development of legislation in Australia regulating artificial intelligence in Australia. For convenience, those provisions in the draft EU AI Act provide that:

    Article 5 – Respect for human rights and fundamental freedoms

Each Party shall take the necessary measures to ensure that all activities in relation to the design, development, use and decommissioning of artificial intelligence systems are compatible with relevant human rights and non-discrimination obligations undertaken by it under international law, or prescribed by its domestic law.

    Article 6 – Integrity of democratic processes and respect for rule of law

1. Each Party shall take the necessary measures to protect the ability of anyone to reach informed decisions free from undue influence or manipulation through the use of artificial intelligence systems in the context of equal and fair access to public debate and democratic processes, in particular democratic participation, freedom of assembly and freedom of expression, including the freedom to seek, receive and impart information of all kinds, as well the freedom to hold opinions without interference.

2. Each Party shall take the necessary measures to ensure that artificial intelligence systems are not used to undermine the integrity, independence and effectiveness of democratic institutions and processes, including respect for judicial independence and the principle of separation of powers.

7.    In our submission, these provisions align with the requirements pursuant to Australia’s obligations under the OECD AI Principles that requires:

    1.2. Human-centred values and fairness

a) AI actors should respect the rule of law, human rights and democratic values, throughout the AI system lifecycle. These include freedom, dignity and autonomy, privacy and data protection, non-discrimination and equality, diversity, fairness, social justice, and internationally recognised labour rights.

b) To this end, AI actors should implement mechanisms and safeguards, such as capacity for human determination, that are appropriate to the context and consistent with the state of art.[4]

8.    Furthermore, it is our view that these provisions, and generally the regulation of artificial intelligence including automated decision making (as defined in the Discussion Paper and for which we agree appropriately covers the scope of these computational processes as they currently exist), ought to be contained within enforceable legislation rather than being expressed as principles. It is our view that the legislative framework for artificial intelligence and automated decision making ought to be framed in a way that prohibits the use of these systems unless they comply with a licensing regime that has a primary focus on human rights and human-centric design.

9.    This submission is bolstered by recent research that indicates that “few Australian organisations have implemented systematic and structured forms of governance around AI systems”[5] and further supported by the concerns highlighted in the Report of the Royal Commission into the Robodebt Scheme[6]. It is relevant and important that these findings have occurred despite Australia’s introduction of Australia’s principle-based AI Ethics Framework in 2019.

10.  A principle-based regulatory framework for regulating artificial intelligence is unlikely to achieve the level of protection of Australians that would be achieved if the regulation of artificial intelligence and automated decision making occurs by legislative force. This requires an Act of Parliament that prohibits the development or application of these systems other than where they have been assessed to comply with a license or authorisation that has human rights as a primary consideration.

11.  We trust that these submissions assist the consultation and Discussion Paper and we confirm that we are willing to assist further with any public hearing(s) associated with this process.

12.  Please do not hesitate to contact us should you require any further information.

Your sincerely,

Angus Murray, Vice-President

For and on behalf of the Queensland Council for Civil Liberties

[1] LibertyWorks Inc v Commonwealth of Australia [2021] HCA 18 at [298] per Steward J citing Theophanous v Herald & Weekly Times Ltd [1994] HCA 46; 182 CLR 104

[2] Note: Australian Government, ‘Australia’s Human Rights Framework’,  https://www.aph.gov.au/-/media/Committees/Senate/committee/humanrights_ctte/Aus_Human_Rights_Framework/Aust_HR_Framework_2010.pdf?la=en&hash=E28A006D823EE0BCDDCED2C0B851C4E56B4EEE04, accessed 2 June 2023. 

[3] https://www.coe.int/en/web/artificial-intelligence/cai.

[4] Organisation for Economic Co-operaton and Development, ‘Recommendation of the Council of Artificial Intelligence’, OECD Legal Instruments (Recommendation, 22 May 2019) <https://legalinstruments.oecd.org/en/instruments/OECD-LEGAL-0449#adherents>.

[5] Solomon, L., & Davis, N., (2023) The State of AI Governance in Australia, Human Technology Institute, The University of Technology Sydney

[6] Catherine Holmes AC SC, Royal Commission into the Robodebt Scheme (Final Report, Volume 1, 7 July 2023).