Searchlight AI
Australia

In Australia, it is unlawful to discriminate on the basis of a number of protected attributes including age, disability, race, sex, intersex status, gender identity and sexual orientation in certain areas of public life, pursuant to federal legislation including the Age Discrimination Act 2004 (Cth), Disability Discrimination Act 1992 (Cth), Racial Discrimination Act 1975 (Cth), Sex Discrimination Act 1984 (Cth) and Australian Human Rights Commission Act 1986 (Cth).

Under the Fair Work Act 2009 (Cth), if an employee can show that they possess a relevant attribute and that adverse action was taken against them, the burden shifts to the employer to prove that no unlawful bias or discrimination played a role in the decision to take the action. The risk of discrimination may be heightened in circumstances where an AI system has been used to assist with decision making, by virtue of algorithmic bias resulting from an overrepresentation or underrepresentation of certain demographics in the data used to train the system, and this reverse onus of proof may be difficult to discharge due to the “black box” nature of many advanced AI systems.

In December 2023, the Australian Government established a Copyright and Artificial Intelligence Reference Group to consider the copyright challenges emerging from AI. In September 2024, the Government published a report setting out the findings from the Group’s first information gathering exercise relating to the use of copyright material as inputs into AI systems to help inform the Government’s policy response to AI.

While existing intellectual property laws may apply in AI-related contexts, there are some limits to their application. For example:

• It is possible that copyright protections may apply to AI-related works under the Copyright Act 1968 (Cth). However, certain criteria must be met, such as the requirement for the work to be an “original” creation of a qualified person. In the case of AI-generated works, it may be challenging to establish the level of independent intellectual effort required to meet this criterion. A ‘qualified person’ under the Copyright Act is an Australian citizen or person resident in Australia (but can be extended to other countries in some circumstances). This will exclude AI-generated content.
• It is possible that the use of an AI system could give rise to copyright infringement, for example by reproducing a substantial part of a copyright-protected work either as part of the training of the system, or by the AI-generated output.
• The Full Court of the Federal Court has held only a natural person can be named as an inventor for the purposes of a patent application under the Patents Act 1990 (Cth).

Read our article on recent updates and developments relating to IP in AI here.

The Australian Competition and Consumer Commission (ACCC) – Australia’s national competition, consumer, fair trading and product safety regulator – confirmed in its Corporate Plan for 2024-25 that responding effectively to changes in the digital economy – including the risk of AI – remains a central focus.

On 15 October 2024, the Australian Government announced a review of the Australian Consumer Law (ACL) component of the Competition and Consumer Act 2010 (Cth) to explore the application of the ACL in relation to AI-enabled goods and services, which it defines as goods and services which involve a consumer directly interacting with an AI system.

The ACL is a principles-based framework and, while it may apply to AI-enabled goods and services in its current form, the Government identifies certain challenges that may apply in the application of the ACL in such contexts and seeks views on whether the ACL remains suitable to protect consumers who use AI and support the safe and responsible use of AI by businesses.

On 15 November 2024, the Australian Government released a further consultation paper seeking feedback on its proposal to incorporate a general prohibition relating to unfair trading practices into the ACL. The general prohibition would capture conduct that “unreasonably distorts or manipulates, or is likely to unreasonably distort or manipulate, the economic decision-making or behaviour of a consumer, and causes, or is likely to cause, material detriment (financial or otherwise) to the consumer”. The Government notes in its consultation paper that these ‘deceptive patterns’ often occur on a large scale using modern technology (such as AI-enabled tools). 

Both consultations processes are now closed but the Government is yet to publish its response.

Read our article on the ACL reform initiatives here.

In August 2023, Australia’s eSafety Commissioner published a statement setting out its position in relation to Generative AI, including regulatory challenges and approaches and providing advice to users. More recently, the eSafety Commissioner published a submission which broadly supported the Government’s proposed mandatory guidelines for high-risk AI and highlighted some examples of existing guardrails within eSafety’s regulatory powers such as industry codes and standards registered under the Online Safety Act 2021 (Cth).

The eSafety Commissioner also supported the Criminal Code Amendment (Deepfake Sexual Material) Act 2024 which passed in August 2024 to amend the criminal code to impose serious criminal penalties on those who share sexually explicit material without consent. This includes material that is digitally created using artificial intelligence or other technology.

The Australian Communications and Media Authority (ACMA) – the authority responsible for the regulation of broadcasting, radiocommunications and telecommunications in Australia – also published a submission broadly in support of the Government’s proposed mandatory guardrails for high-risk AI, but makes clear in the submissions that many of the guardrails set out the types of actions that the ACMA already expects its regulated entities to be taking pursuant to legislation including the Broadcasting Services Act 1992, the Radiocommunications Act 1992, the Telecommunications Act 1997 and the Spam Act 2003.

The Australian Securities and Investments Commission – Australia's integrated corporate, markets, financial services and consumer credit regulator – has emphasised, including by way of a Report published in October 2024 and comments made by ASIC Chair Joe Longo, that many of the existing obligations that apply to the entities that it regulates are technology neutral such that entities need to ensure their use of AI does not breach any of these provisions. This includes, for example:

• Obligations on Australian financial services licensees under the Corporations Act 2010 (Cth) and on Australian credit licensees pursuant to the National Consumer Credit Protection Act 2009 (Cth), including the obligation to provide services efficiently, honestly and fairly and have adequate systems in place to ensure compliance.

Directors’ duties under the Corporations Act 2010 (Cth), including the duty to act in good faith and with a reasonable degree of care and diligence. The obligation not to engage in misleading or deceptive conduct or make false or misleading representations as prohibited by the Corporations Act 2010 (Cth) and Australia Securities and Investments Commissions Act 2001 (Cth), is relevant where an entity engages in “AI washing” (for example, by overstating their use of AI).

Court protocols and judicial guidelines have been released in various Australian jurisdictions providing guidance on the responsible use of generative AI in court and tribunal proceedings. While the guidelines vary by jurisdiction, they all highlight risks such as inaccuracies resulting from 'hallucinations' produced by AI systems and data privacy concerns, and warn against the improper use of such systems in proceedings.

• The Supreme Court of New South Wales Practice Note, which has been adopted by several other Courts in New South Wales, outlines acceptable uses of generative AI by litigants and judges and prohibits the use of generative AI in certain circumstances including in generating the content of affidavits, witness statements and character references.
• The Guidelines published by the Supreme Court of Victoria, and adopted by the County Court, set out various principles for the use of AI tools by litigants encouraging transparency where AI tools are used and make clear the circumstances in which such tools may be used by the judiciary.
• The Guidelines published by the Queensland Courts were developed to assist non-lawyers who represent themselves or others in court and tribunal proceedings and encourage users to be aware of ethical issues associated with the use of generative AI tools.

• Bar associations and law societies in various Australian jurisdictions have also issued guidance notes.