← All articles

Data Breach Penalties in Australia: What Your Business Risks

What a data breach can cost an Australian business: penalties up to $50 million, infringement notices, real cases, and steps to cut your exposure.

Data Breach Penalties in Australia: What Your Business Risks

An Australian organisation that suffers a serious data breach can face Federal Court civil penalties of up to $50 million, three times the benefit it gained, or 30% of its adjusted turnover per contravention. Smaller infringement notices apply for process failures such as not notifying a breach. For contraventions after 13 December 2022, the old ceiling of roughly $2.22 million no longer applies. This guide sets out which penalties apply, what the first court outcome looked like, and what you can do about it before a breach happens.

TL;DR

  • Serious or repeated interferences with privacy now carry maximum penalties of the greater of $50 million, three times the benefit obtained, or 30% of adjusted turnover, per contravention, as described by the OAIC.
  • The first civil penalty under the Privacy Act was $5.8 million against Australian Clinical Labs, ordered in October 2025 under the older, lower regime.
  • Failing to notify an eligible data breach can draw an infringement notice, with no court case needed.
  • The 30-day assessment window is a ceiling. Regulators expect faster action, and a documented response plan is your best defence.
  • Where your data sits, and who can touch it, shapes both the likelihood and the severity of a breach.

Table of Contents

The penalty regime in one view

The Privacy Act 1988 (Cth) does not set one fine for “a data breach”. A breach by itself is not automatically a penalty event. Penalties follow from what the organisation did or failed to do. In practice, three tracks matter.

TrackTriggerWho decidesMaximum exposure
Serious or repeated interference with privacyFailing to take reasonable steps to protect personal information, with serious consequencesFederal Court, on application by the CommissionerGreater of $50 million, three times the benefit, or 30% of adjusted turnover, per contravention (conduct after 13 December 2022)
Infringement noticesSpecific process failures, such as not notifying an eligible data breachThe OAIC directly60 penalty units ($21,840) for a body corporate; 200 penalty units ($72,800) for certain listed corporations
Other regulatory actionInvestigations, determinations, enforceable undertakingsThe CommissionerCompensation orders, undertakings, public findings

The figures for the first two rows come from the OAIC’s penalty announcement and its infringement notices page. The Commissioner’s power to ask the court for orders sits in section 80W of the Act, explained in the OAIC’s guide to civil penalties.

Two points catch people out. First, “per contravention” matters. In its civil penalty action against Optus, the regulator alleges one contravention for each of the roughly 9.5 million individuals affected. The count of people, not the count of incidents, can drive the arithmetic. Second, the higher ceilings did not apply to Optus because the conduct predated them. Conduct from December 2022 onwards is a different proposition.

What changed in December 2022

Before 13 December 2022, the maximum for a serious interference with privacy was about $2.22 million per contravention. After it, the ceiling became the greater of three figures: $50 million, three times the benefit derived from the conduct, or 30% of adjusted turnover. For a large enterprise, the turnover test is the one to model. A business with $1 billion in adjusted turnover is looking at a theoretical ceiling far above $50 million.

Comparison of Australian privacy penalty ceilings

Maximums are not typical outcomes. The court decides the figure, weighing the seriousness of the conduct, the harm, and what the organisation did afterwards. But a ceiling that high changes how a board should treat the risk. It stops being a compliance line item and becomes a financial exposure to be reported, tested and insured against.

The OAIC has been clear about where this leads. Its Privacy Commissioner, Carly Kind, said the first penalty outcome was a turning point in enforcement, and singled out healthcare providers in particular.

The first civil penalty: Australian Clinical Labs

In October 2025 the Federal Court ordered Australian Clinical Labs to pay $5.8 million over a February 2022 cyber-attack on its Medlab Pathology business. These were the first civil penalties ever ordered under the Privacy Act, and the OAIC’s summary of the case breaks the amount into three parts:

  1. $4.2 million for failing to take reasonable steps to protect personal information (Australian Privacy Principle 11.1).
  2. $0.8 million for failing to carry out a reasonable assessment of whether the incident was an eligible data breach.
  3. $0.8 million for failing to give the Commissioner the required statement.

The lesson is that two of the three penalties were about response, not prevention. The organisation was penalised for how it assessed and reported the incident, not only for the security gap itself. Because the conduct happened before December 2022, the older maximum of $2.22 million per contravention applied. The same conduct today would sit under the far higher regime.

Infringement notices: penalties without a trial

Not every consequence goes to court. The OAIC can issue an infringement notice as an alternative to litigation for specific contraventions. These include failing to notify an eligible data breach, failing to keep a compliant privacy policy, and failing to give effect to an opt-out request.

Per the OAIC, a single notice can reach 60 penalty units ($21,840) for a body corporate, and 200 penalty units ($72,800) for certain listed corporations. Those sums look modest next to $50 million. They matter for two reasons. They are quick, because no court is involved. And a pattern of small notices tells the regulator something about your culture, which can shape whether it escalates.

When is a breach “notifiable”?

The Notifiable Data Breaches scheme applies when personal information is lost or accessed or disclosed without authorisation, the incident is likely to result in serious harm to one or more individuals, and the organisation has not been able to prevent that harm through remedial action. The OAIC sets this out on its data breach dashboard announcement.

If you suspect a breach, the law requires you to assess it within 30 days. The OAIC treats that as a maximum, not a target. Its guidance on the NDB scheme expects assessments to be done expeditiously, and it expects you to be able to document why if you run past 30 days. Waiting for executive sign-off is not a recognised reason to delay. The clock is tied to when your people have reasonable grounds to suspect, not when leadership is ready.

If the breach is eligible, you notify the Commissioner and the affected individuals. The OAIC’s quick reference guide is a useful one-page check for your incident team. For a deeper walk-through aimed at compliance staff, see our guide to notifiable data breaches for Australian compliance teams.

When is a breach "notifiable"? — overview diagram

Why the risk is rising

The OAIC received a record 1,205 data breach notifications in 2025, up 8% on 2024, according to its annual statistics release. Malicious or criminal activity caused the majority, with 716 notifications. Health and finance were the two most affected sectors, at 19% and 15% of notifications.

Privacy Commissioner Carly Kind has described the threat as substantial and rising year on year. For IT and operations managers in health, finance and professional services, that means more scrutiny and a higher bar for auditable evidence of how you handle data and respond to incidents.

The costs that never appear in the penalty table

A penalty figure understates the bill. Add the following:

  • Investigation and legal costs. Engaging forensic specialists and counsel, often for months.
  • Notification and remediation. Contacting every affected person, running a call centre, offering credit monitoring or ID support.
  • Compensation. The Commissioner can seek compensation for affected individuals, and class actions are a separate risk.
  • Customer churn and reputational damage. Harder to model, usually the longest-lasting cost.
  • Management time. Executives and IT teams are diverted from normal work for a long stretch.

We have collected real-world examples in our article on data breach consequences and what to do next.

How to reduce your penalty exposure

Penalties turn on whether you took reasonable steps. That is a practical test, and you can prepare for it. A workable order of work looks like this:

  1. Map your personal information. List what you hold, where it lives, who can access it, and which vendors process it. You cannot protect or assess what you cannot find.
  2. Write and rehearse an incident response plan. Name who decides, who assesses, who notifies and who talks to the media. Run a tabletop exercise at least once a year. If you need outside help building one, a managed IT provider such as Secure Techies offers incident response and risk assessment services, though its pages are written for a US audience, so check them against Australian requirements.
  3. Start the assessment clock early. Define internally what counts as “reasonable grounds to suspect” and who must escalate. Do not wait for certainty.
  4. Keep records. If you take longer than a few days, write down why. The OAIC expects you to be able to explain it.
  5. Minimise and localise data. Data you do not collect cannot leak, and data that stays under Australian jurisdiction is simpler to govern. See our checklist on preventing offshore data transfers in AI projects.
  6. Audit your vendors. Many breaches begin with a supplier. Ask where data is stored, who has access, and what certifications they hold. Our 12 questions on Australian data sovereignty is a good starting list, and health providers can use the five vendor checks for hosting health data.
  7. Report upward. Give your board a figure for exposure, not only a heat map. The turnover-based ceiling makes this straightforward to calculate.

Where automated customer communication fits

Contact centres, receptionists and chat channels handle large volumes of names, health details, account numbers and payment conversations. Every one of those interactions is personal information in transit or at rest. If you adopt AI agents to handle them, the questions that matter for penalty exposure are: where is the data hosted, who can access transcripts and recordings, and can you prove it?

For regulated sectors, the OAIC’s expectations on chatbots are worth reading before you deploy. We summarise them in what to do in 48 hours on OAIC chatbot guidance, and cover sector specifics in our guides for healthcare chatbot compliance and secure AI for finance.

Our perspective

We build a conversational AI platform for Australian organisations, so we see the penalty regime from the procurement side. The pattern is consistent. Teams ask about features first and data location last. The penalty framework suggests it should be the other way round.

Here is how we think about it. A breach penalty rests on whether you took reasonable steps to protect personal information. When voice, SMS, email and chat data flow through a vendor, your steps include choosing where that data lives. That is why our platform is hosted entirely within Australia, as a private cloud, with SOC 2 certification and full voice AI and CRM integration in one place. Fewer hops and fewer offshore processors mean fewer places for something to go wrong, and a shorter, clearer answer when a regulator asks where the data went.

For organisations that need more control still, we also offer on-premises deployment. We do not claim a platform removes your obligations. Your assessment, notification and governance duties remain yours. What a well-chosen platform does is shrink the surface you have to defend and make your evidence easier to produce.

Our advice is simple. Put your incident response plan on paper before you buy anything. Then hold every vendor, including us, to it.

— Sowrabh

Built for regulated Australian teams

If you are automating customer conversations in healthcare, finance, real estate or collections, we can show you how our Australia-hosted platform handles voice, SMS, email and chat with CRM integration. See our conversational AI platform or go straight to the page for banking and finance or healthcare. Book a demo and bring your security questionnaire.

FAQ

What is the maximum penalty for a data breach in Australia?

For serious or repeated interferences with privacy after 13 December 2022, the court can impose up to the greater of $50 million, three times the benefit obtained, or 30% of adjusted turnover, per contravention. Earlier conduct falls under the old cap of about $2.22 million per contravention. The court sets the actual amount.

Is every data breach penalised?

No. A breach alone does not trigger a penalty. Exposure arises where an organisation failed to take reasonable steps to protect information, failed to assess or notify properly, or breached another provision. The first penalty case involved all three of those failures.

How long do I have to assess a suspected breach?

The law allows 30 days, but the OAIC treats that as an outer limit and expects you to act expeditiously. Document the reasons if you take longer. Internal approval delays are not an accepted justification.

Can the OAIC fine a business without going to court?

Yes, for specific contraventions. It can issue infringement notices of up to 60 penalty units ($21,840) for a body corporate, or 200 penalty units ($72,800) for certain listed corporations, for matters such as failing to notify an eligible data breach.

Sources

Jess, AI voice agent