The AHRC's role in AI governance
The Australian Human Rights Commission has positioned itself as a central voice in Australia's AI governance debate. Its 2021 Human Rights and Technology Final Report — produced after a two-year national consultation — remains the most comprehensive analysis of AI's impact on human rights in the Australian context. The Commission's recommendations, while not all adopted as legislation, have significantly shaped government AI policy and provide a framework for organisations assessing the human rights implications of their AI deployments.
The Commission's central argument is that AI governance in Australia must be anchored in human rights — not merely in economic efficiency or innovation policy. This means that AI systems must be assessed against the specific rights they may affect: the right to equality and non-discrimination, the right to privacy, the right to a fair hearing, freedom of expression, and the rights of children. For organisations deploying AI, this framework provides a structured way to assess and document the human rights implications of their systems.
Anti-discrimination law and algorithmic discrimination
Australian federal anti-discrimination legislation — the Age Discrimination Act 2004, the Disability Discrimination Act 1992, the Racial Discrimination Act 1975, and the Sex Discrimination Act 1984 — prohibits both direct and indirect discrimination on the basis of protected attributes. Direct discrimination means treating someone less favourably because of a protected attribute. Indirect discrimination means imposing a condition or requirement that is facially neutral but has a disproportionate adverse effect on people with a protected attribute and that cannot be justified.
Indirect discrimination is the primary discrimination risk from AI. An AI hiring tool that uses educational institution as a variable may indirectly discriminate by race if certain racial groups are underrepresented in particular institutions. An AI credit scoring model that uses suburb or postcode may indirectly discriminate by race if those areas are correlated with racial demographics. An AI health assessment tool trained on data from populations that excluded people with disabilities may perform less accurately for disabled people. In each case, the algorithm is facially neutral — it does not explicitly reference a protected attribute — but it produces outcomes that disadvantage protected groups.
The AHRC complaints pathway
The AHRC can accept complaints under federal anti-discrimination legislation from individuals who believe they have been discriminated against. An individual who believes an AI system discriminated against them — in a hiring process, a credit decision, a service eligibility assessment, or any other context covered by anti-discrimination law — can file a complaint with the Commission. The Commission will attempt conciliation between the complainant and the respondent. If conciliation fails, the complaint can be terminated and the complainant can then take their case to the Federal Court or the Federal Circuit and Family Court.
