Quick answer: Australia has no standalone AI Act yet. Businesses using AI must already follow existing laws, including the Privacy Act, the Australian Consumer Law, anti-discrimination, workplace and copyright law. From 10 December 2026, privacy policies must disclose significant automated decisions. Mandatory Australian Standards for AI are planned for 2027, aimed first at data centres and frontier AI developers.
It's one of the most common questions Australian business owners ask before they switch on a chatbot or an AI agent: are there AI laws in Australia, and do they apply to a business like mine? The short answer is yes, just not in the form most people expect. There's no single AI rulebook. Instead, the laws you already follow apply to AI the moment it touches customers, staff or personal information. If you're planning custom AI solutions, here's what applies today and what's coming in 2027.
Is There an AI Act in Australia?
Not yet. In September 2024, the federal government proposed 10 mandatory guardrails for high-risk AI. In August 2025, the Productivity Commission recommended pausing that work, and on 2 December 2025 the National AI Plan shelved the guardrails in favour of existing technology-neutral laws, voluntary guidance and a new AI Safety Institute.
The direction changed again in 2026. On 15 July 2026, the Prime Minister announced mandatory Australian Standards for AI and set up an Office of AI inside the Department of the Prime Minister and Cabinet. National Cabinet endorsed developing the standards as mandatory on 26 August 2026, and legislation is expected in early 2027.
Here's the part most headlines miss. As currently framed, the first standards target large data centres, frontier AI training and the use of Australian creative works, not ordinary businesses that deploy AI. On 10 November 2026, the government also proposed a dedicated safety regulator for frontier AI companies. For a typical SMB, existing law is still where the real obligations sit.
5 Existing Laws That Apply to AI Right Now
| Law | What it means when you use AI | Regulator |
|---|---|---|
| Privacy Act 1988 | Personal information in prompts, training data and outputs must meet the Australian Privacy Principles. A statutory tort for serious invasions of privacy has applied since June 2025. | OAIC |
| Australian Consumer Law | What a chatbot tells customers about prices, refunds or rights counts as your representation. Overstating your AI capability (AI-washing) is also misleading conduct. | ACCC and state fair trading |
| Anti-discrimination laws | Biased outputs in hiring, lending or pricing can breach federal and state discrimination laws, even if nobody intended it. | Australian Human Rights Commission and state bodies |
| Fair Work Act and modern awards | Most awards require consultation before major workplace change, including new technology. Some states also regulate workplace surveillance. | Fair Work Commission and Fair Work Ombudsman |
| Copyright Act 1968 | Training data and AI-generated content raise ownership questions. The government has ruled out a text and data mining exception. | Courts |
The privacy change is the one with a hard date. From 10 December 2026, organisations covered by the Privacy Act must explain in their privacy policy when a computer program uses personal information to make, or substantially help make, decisions that significantly affect people. The OAIC's guidance shows this reaches well beyond obvious AI, including workflows and even spreadsheets with decision formulas.
Consumer law carries the biggest penalties. For companies, breaches can cost the greater of $50 million, three times the benefit gained, or 30% of adjusted turnover for the breach period.
What Australia's AI Guidance Asks Businesses to Do
On 21 October 2025, the National AI Centre released the Guidance for AI Adoption, which replaced the 2024 Voluntary AI Safety Standard. It's voluntary, but regulators and enterprise buyers increasingly treat it as the benchmark. It condenses the old 10 guardrails into six practices:
- Decide who is accountable. Name an owner for each AI system and for AI use overall.
- Understand impacts and plan accordingly. Work out who could be affected before you deploy.
- Measure and manage risks. Add AI to your existing risk register, not a separate silo.
- Share essential information. Tell people when and how AI is used.
- Test and monitor. Check accuracy before launch and keep checking after.
- Maintain human control. Keep a person able to review, override or switch off the system.
The guidance includes a free AI policy template, an AI register template and a screening tool. A separate draft workplace AI guide was released on 12 November 2026, with public comment open until 27 January 2027.
Extra Rules for Health, Finance and Other Regulated Sectors
Some industries carry an extra layer on top of the general laws. If you operate in one of these, check the sector rules first.
- Healthcare: AI that diagnoses, monitors or recommends treatment can count as a medical device regulated by the TGA. Practitioners also have AHPRA guidance on using AI in care.
- Banking, insurance and super: APRA's CPS 230 (operational risk) and CPS 234 (information security) apply to AI systems and the vendors behind them.
- Financial advice and credit: ASIC expects licensees to govern AI under their existing obligations, including fair treatment and record-keeping.
- Legal services: Courts and law societies have issued rules on generative AI in court documents and client work.
- Government suppliers: Agencies follow the federal policy for responsible AI use, and those expectations flow into supplier contracts.
For everyone else, the five laws in the table above are the baseline.
An Australian Example: AI CV Screening in a Recruitment Firm
Consider a Melbourne recruitment agency with $5 million in turnover that uses an AI tool to rank job applicants. It's above the Privacy Act's small business threshold, so the Australian Privacy Principles apply to every CV it processes.
Because a shortlist decision significantly affects a candidate, the agency's privacy policy must describe that automated decision-making from 10 December 2026. If the tool keeps ranking older applicants lower, the agency, not the software vendor, faces the discrimination complaint. One law doesn't cover this scenario. Three do.
What to Watch in 2027
The rules will keep moving next year, so it's worth knowing which dates matter. The main ones to track:
- Australian Standards for AI legislation, expected in Parliament in early 2027. Watch whether any obligations extend to businesses that deploy AI.
- The frontier AI safety regulator proposed on 10 November 2026, and its penalty powers.
- The Joint Select Committee on AI report, due 30 November 2026, which may shape the bill.
- The next round of Privacy Act reforms, including a possible end to the small business exemption.
- The final workplace AI guide, after consultation closes on 27 January 2027.
The smartest move now isn't to wait for the bill. Get your AI register, privacy policy and staff policy in order under today's laws, and the 2027 changes become an update rather than a rebuild.
Plan your next steps with Zynex Technologies
Discuss your AI security and compliance needs
Talk to the Zynex Technologies team about your business and its AI systems.
Book a call