AI in Australian HR and Employment: What People Teams Must Get Right

The AI tools HR teams are using — and the obligations they create

Australian HR and people teams are now using AI across the employment lifecycle: applicant tracking systems with AI screening; video interview platforms that assess candidate responses, facial expressions, and voice patterns; AI-driven reference checking; psychometric assessment AI; performance management systems that generate automated scores from activity data; workforce analytics that identify patterns in attendance, productivity, and engagement; AI-powered scheduling and rostering; and employee survey analysis AI. Each of these creates specific legal obligations that HR professionals need to understand, because the employment law consequences of getting it wrong are significant.

AI in recruitment: anti-discrimination is your primary risk

Australian anti-discrimination law operates at both federal and state levels, and collectively prohibits discrimination on the basis of: sex, sexual orientation, marital status, pregnancy, and family responsibilities; race, colour, national origin, and ethnic background; disability and medical condition; age; religious belief; and political opinion. These protections apply in employment, including in the recruitment process.

AI recruitment tools create indirect discrimination risk through proxy variables. A tool trained on historical hiring data may learn to associate success with characteristics that correlate with protected attributes without explicitly encoding those attributes. The result is systematic discrimination against protected groups that operates at a scale and speed impossible with human recruiters — and that may not be visible in aggregate hiring statistics until patterns emerge across many hiring decisions.

Governance requirements for recruitment AI: independent bias testing across all relevant protected attributes before deployment; ongoing monitoring of selection rates across demographic groups; human review of all AI-recommended shortlists before final decisions; and candidate disclosure that AI was used in their assessment. The NSW Anti-Discrimination Act, Sex Discrimination Act, and Disability Discrimination Act all apply regardless of whether discrimination is intentional or algorithmic.

Performance management AI and procedural fairness

AI-generated performance data — productivity scores, activity metrics, quality assessments — is increasingly used in performance management and disciplinary processes. Under the Fair Work Act, employees are entitled to procedural fairness in performance management and disciplinary processes. This means: being informed of the specific performance concern in terms they can meaningfully respond to; having an opportunity to respond; and having that response genuinely considered.

"Your AI productivity score was below threshold" is not procedural fairness. Employees have a right to understand what behaviours or outputs contributed to that score, why those factors are relevant to their role, and how to demonstrate improvement. HR teams using AI performance systems should ensure the systems can produce explainable outputs that support meaningful performance conversations — not just summary scores.

Workplace monitoring: state law variation

The legal framework for AI-powered employee monitoring varies significantly by state. In New South Wales, the Workplace Surveillance Act 2005 requires 14 days written notice before any surveillance system is introduced, and prohibits covert surveillance without specific authority. The Act covers computer surveillance (which captures AI monitoring of computer activity), camera surveillance, and tracking. In the Australian Capital Territory, the Workplace Privacy Act 2011 has similar provisions. In all other states and territories, the Privacy Act's Australian Privacy Principles apply to employee personal information, along with general employment law constraints on unreasonable monitoring. The absence of specific surveillance legislation in Victoria, Queensland, South Australia, Western Australia, Tasmania, and the Northern Territory does not mean monitoring is unregulated — it means the regulatory framework is less specific.