Victoria Pledges Rules for AI Hiring and Worker Surveillance

Victoria’s Labor government pledged on July 20 to introduce workplace-surveillance rules that would restrict biometric and AI monitoring, require notice and consultation, and protect hiring and pay decisions from discrimination. The proposal would also require human review of significant automated decisions and let workers access surveillance data used against them. The measures are an election commitment, not enacted law; Labor says it would legislate them if returned in November.
Victoria’s Labor government pledged on July 20, 2026, to introduce what it calls Australia’s toughest workplace-surveillance protections if it is returned at the state election in November. The proposal covers employee monitoring, biometric data and AI-assisted decisions in recruitment, promotion and pay.
What Victoria proposed
The government’s announcement says employers would have to tell workers when they are being monitored and consult them before introducing surveillance technology or changing how it is used. Surveillance would need a legitimate purpose, while covert monitoring would be restricted to cases involving suspected unlawful activity.
The plan would also limit collection of biometric data to situations where no less intrusive option exists. Using biometrics to infer emotions would be prohibited unless there is a legitimate reason, such as fatigue monitoring for truck drivers. The government also singled out AI systems that track bathroom breaks, pregnancy or physical characteristics without a specific purpose.
For significant automated decisions based on surveillance data, a human would have to review the outcome before it affects a worker’s rights. Workers would also gain access to surveillance data used in employment decisions.
Hiring rules are part of the same pledge
The proposal extends beyond monitoring. Victorian Labor says AI used in hiring, promotion or pay must not discriminate on characteristics such as race, gender, sexuality or religion. Human Resources Director Australia reported that the government plans to amend the state’s equal-opportunity framework as part of the package.
The policy is not yet law. No enacted bill, final statutory language or commencement date accompanied the announcement, so the eventual scope and enforcement mechanism will depend on legislation after the election.
What AI and data teams should watch
For employers and vendors, the practical signal is that model governance may need to cover more than accuracy. If the pledge becomes law, organisations operating in Victoria could need documented purposes for monitoring, limits on biometric collection, bias controls for employment models, consistent human-review procedures and a way for workers to inspect data used in consequential decisions.
The originating announcement establishes the policy commitment. Independent HR reporting confirms its election-stage status and recruitment focus, while The Conversation’s expert analysis examines why clearer rules could also reduce uncertainty for employers. The next decisive event will be publication of draft legislation, which should reveal definitions, exemptions, enforcement powers and the evidence required to demonstrate compliance.
Key Points
- 1Victoria’s Labor government promised new rules covering AI-assisted hiring, biometric monitoring and workplace surveillance if it wins the November election.
- 2The proposal would require worker notice and consultation, human review of significant automated decisions, and access to surveillance data used in employment outcomes.
- 3The measures remain a policy pledge rather than enacted law; final scope, exemptions and enforcement will depend on future legislation.
Scoring Rationale
The pledge could create material governance obligations for AI-assisted hiring and worker-monitoring systems in Victoria. Its reach is regional and the rules are not yet enacted, which limits immediate operational impact.
Sources
Primary source and supporting public references used for this report.
Practice interview problems based on real data
1,625 SQL & Python problems across 15 industry datasets — the exact type of data you work with.
Try 250 free problems


