Data Breach Response Plan: Assessment and Notification
Skip to content

Data Breach Response Plan

A data breach response plan sets out how an organization will respond when personal or sensitive information is accessed, disclosed, or lost without authorization. It covers containing the breach, assessing the harm, notifying those affected, and preventing a repeat.

The hardest part of any breach is deciding whether it must be notified. This guide works through the timeline and the test, then applies them to worked examples.

How Data Breaches Happen

A data breach occurs when information an organization holds is accessed or disclosed without authorization, or is lost in circumstances where unauthorized access or disclosure is likely. Regulators group the causes into malicious or criminal attack, human error, and system fault, and a plan needs to handle each.

Malicious or Criminal Attack

Malicious or criminal attack is the largest category in the Notifiable Data Breaches reports published by the Office of the Australian Information Commissioner. It includes phishing that leads to a compromised mailbox, stolen or guessed credentials, ransomware with data theft, and intrusions through unpatched systems. It also includes theft of paperwork and devices, and deliberate action by an insider.

These breaches are the hardest to assess, because the organization usually does not know at first what the attacker reached.

Human Error

Human error covers personal information sent to the wrong recipient by email, post, or message, files published or shared with the wrong settings, a reply-all that should have been blind copied, and documents left in a public place. These breaches are frequent, usually small, and often containable if caught quickly.

System Fault

A system fault is a software bug or misconfiguration that exposes records to the wrong users: a portal that shows one customer another’s details, or a database reachable without a password.

A breach does not need an attacker, so the plan has to be usable by an office manager dealing with a misdirected email as well as by a security team dealing with ransomware. Many breaches also involve email in some way, either as the route in or as the place the data was sitting.

A cyber incident and a data breach are different things that often overlap. An incident response plan deals with the attacker and the systems. A data breach response plan deals with the information and the people it belongs to. The two run side by side.

Containing a Breach

Containment is the first duty and the one that most affects the outcome. Effective remedial action can change whether a breach is notifiable at all.

Misdirected email

Recall it if the mail system allows. Contact the recipient at once, ask them to delete it unopened, and get written confirmation. Note who the recipient is and whether they can be trusted.

Wrongly shared file or link

Remove the sharing, then check the access log to see who opened it while it was exposed.

Compromised mailbox or account

Preserve the audit records first, then run the compromised account playbook set out in the incident response playbooks guide.

Lost or stolen device

Confirm from the device management record whether it was encrypted and locked, then run the lost device playbook in the same guide.

Exposed system or database

Restrict access at once, then copy its access logs before anything else on the system is changed.

Ransomware with data theft

Follow the incident response plan. Containment here is a security operation.

Supplier breach

Ask the supplier what was affected, what they have done, and for their evidence. Review what data they hold.

In every case, record what was done and when. The record of containment is part of the evidence for the assessment that follows.

The Clock Under Australian Law

Australia’s Notifiable Data Breaches scheme operates under the Privacy Act 1988 and is administered by the Office of the Australian Information Commissioner (OAIC). It sets a sequence that a plan should follow.

Point in timeWhat the organization must do
AwarenessStaff notice or are told of a possible breach. Contain it at once: stop the access, recall the message, remove the link.
Grounds to suspectWhere there are reasonable grounds to suspect an eligible data breach, start an assessment. It must be reasonable and expeditious.
Within 30 daysTake all reasonable steps to complete the assessment within 30 days of becoming aware of the grounds to suspect.
Grounds to believeOnce there are reasonable grounds to believe an eligible data breach has occurred, prepare a statement for the OAIC.
As soon as practicableGive the statement to the OAIC, and notify affected individuals.
AfterwardReview the cause, fix it, and record the breach and the decisions made.

The 30 days is an outer limit for the assessment, not a waiting period. An organization that already has enough information to believe a breach is eligible should notify as soon as practicable.

The scheme applies to organizations and agencies covered by the Privacy Act. That generally includes businesses with annual turnover above AUD 3 million, Australian Government agencies, and some smaller businesses such as health service providers.

The Eligible Data Breach Test

A data breach is eligible only when each of the following is true.

1
A breach occurredThere was unauthorized access to, or unauthorized disclosure of, personal information, or a loss where either is likely.
2
Serious harm is likelyA reasonable person would conclude the breach is likely to result in serious harm to any individual the information relates to.
3
Remedial action has not prevented itThe organization has been unable to remove the likely risk of serious harm through the steps it took.

Serious harm is wider than financial loss. It includes identity theft, physical harm, psychological distress, and damage to reputation.

The assessment of likelihood weighs several factors:

The kind of information

Identity documents, health records, financial details, and credentials carry the greatest risk.

Its sensitivity

Information that could embarrass, endanger, or discriminate against a person.

Whether it was protected

Encryption, and whether the key was also exposed.

Who obtained it

A criminal group, an unknown party, or a known recipient who can be trusted to delete it.

How many people

The scale raises the chance that someone is seriously harmed.

How long it was exposed

A longer window widens what could have been taken.

What harm could follow

Fraud, impersonation, extortion, or targeting of vulnerable people.

Record how each factor was weighed.

Worked Assessments

A Payroll File Sent to the Wrong Person

A payroll officer emails a spreadsheet of staff salaries and bank details to an external accountant with a similar name to the intended recipient. The mistake is noticed within the hour.

The recipient is a known professional, confirms in writing that the file was deleted unopened, and has no reason to misuse it. Remedial action has probably removed the likely risk of serious harm. This would usually be assessed as not eligible. It is still a breach, and it goes in the breach register with the reasoning.

A Stolen Laptop

A manager’s laptop is stolen from a car. It holds customer contracts and contact details.

The laptop uses full-disk encryption, the user was signed out, and the device management system shows it has not connected since the theft. Access to the information is unlikely. With evidence of encryption recorded, this would usually be assessed as not eligible. Without encryption, the same theft would probably be notifiable. The lost device playbook in the incident response playbooks guide covers the steps that produce that evidence.

A Mailbox Takeover

An HR coordinator’s mailbox is accessed by an attacker for nine days. Audit records show the mailbox was synchronized in full. It contains job applications with passport scans, tax file numbers, and medical certificates.

Identity documents and health information were obtained by a criminal, for hundreds of people, and nothing done afterward can retrieve them. Every element of the test is met. This is an eligible data breach. The OAIC and the affected individuals are notified as soon as practicable.

A Breach at a Supplier

A payroll provider tells its customers that an attacker copied files from its systems. The files include the names, bank details, and tax file numbers of the organization’s staff.

Both organizations hold the information, so both have obligations under the scheme. The organization’s own assessment starts when the supplier’s notice gives it grounds to suspect, and it depends on the supplier for evidence of what was taken. Bank details and tax file numbers in criminal hands make serious harm likely, and neither party can retrieve them. This is an eligible data breach. Only one of the organizations needs to notify, and OAIC guidance suggests it be the one with the most direct relationship with the individuals. Here that is the employer.

Each case turns on evidence. Proof of deletion, proof of encryption, audit records of access, and a supplier’s account of what was copied decided the outcome. A plan that ensures such evidence exists is worth more than one that only describes the steps.

Running the Assessment

An assessment has to produce a decision that can be defended later. In practice it runs as a short investigation with a fixed set of questions.

Establish the Facts

Set down what happened, when, and how it was discovered, which systems, accounts, or documents were involved, and whether the exposure has stopped.

Identify the Information

This is usually the slowest step. For a misdirected file it takes minutes. For a compromised mailbox holding years of correspondence, it can take weeks.

The work is to find every piece of personal information in what was exposed. For mailboxes and file shares this means searching the content for identity documents, tax and health identifiers, bank details, and dates of birth, then building a list of the people concerned. Specialist review tools exist for this, and for large volumes a forensic or legal review provider is normally engaged.

Establish What Was Accessed

Use audit records to separate what the attacker opened from what they could merely reach. Where a mailbox was synchronized or a share copied in bulk, treat all of it as taken. Where logs are missing, say so, and assess on the basis of what was accessible. The data exfiltration investigation guide covers the evidence in detail.

Assess the Harm

Apply the factors set out earlier to each group of affected people. The answer may differ between groups. Staff whose payroll details were exposed face a different risk from customers whose names and email addresses were.

Decide and Record

Write down the conclusion, the reasoning, the evidence relied on, and who made the decision. Do this whether the outcome is to notify or not.

Work in Parallel

The 30 days pass quickly when content review is involved. Start the identification of personal information on day one, alongside the technical investigation, and prepare notification materials before the assessment concludes.

Writing the Notification

A notification to the OAIC must include the organization’s identity and contact details, a description of the breach, the kinds of information concerned, and recommended steps for individuals. A notice to affected people should cover the same ground in plain words.

What happened

A short factual account, with dates.

What information was involved

The specific types, for that person where possible.

What has been done

Containment, and the steps taken to reduce harm.

What the person can do

Practical actions, such as replacing a document, changing passwords, or watching for scams.

Where to get help

The phone line or mailbox set up to answer questions.

Avoid statements the evidence cannot support. A notice that has to be corrected later damages trust more than an early notice that is honest about what is still unknown.

The scheme gives a choice of route. An organization can notify every individual whose information was involved, or only those at risk of serious harm. Where neither is practicable, it publishes the statement on its website and takes reasonable steps to publicize it.

Notifying at Scale

Notifying a handful of people is a letter. Notifying thousands is a project, and plans that stop at “notify affected individuals” underestimate it.

Contact details

The organization needs a current address, email, or phone number for each person. Former customers and past job applicants are often the hardest to reach.

The channel

Email is fastest. Where the breach involved the email system, or where recipients may suspect a scam, post or a notice on the website may be needed as well.

A place to ask questions

A staffed phone line or mailbox, with people briefed on what they can say. Call volumes peak in the first days.

Help that matches the harm

Guidance on replacing identity documents, access to identity or credit monitoring where appropriate, and referral to support services.

Scam warnings

Criminals follow breach announcements with phishing that pretends to come from the organization. Tell people how genuine contact will look.

Staff briefing

Employees will be asked by customers, family, and sometimes media. Give them a short, accurate script.

Coordination

Legal counsel and the insurer should see the notice before it goes, and the statement to the regulator is lodged first or at the same time. Where several organizations hold the same data, agree who notifies so people are not told twice or not at all.

After Notification

The work continues after the notices are sent.

Regulator follow-up

The regulator may ask how the breach occurred, what was done, and what has changed. The assessment record and incident log are the basis for the answers.

Complaints and claims

Individuals may complain to the organization or the regulator, or seek compensation. Keep records of every contact.

The review

Identify the root cause and the changes needed. For a misdirected email that might be a delay-send rule and a check on external recipients. For a mailbox takeover it is stronger authentication and less data kept in mail.

Data minimization

Many serious breaches involve information that no longer needed to be held. Review what is kept, where, and for how long. Old identity documents in mailboxes are a common example.

What the Plan Should Contain

A plain definition of a breach

With examples, so staff recognize one and report it.

One reporting route

Where to report a suspected breach, and how fast it is acknowledged.

The response team

Named roles, deputies, and contact details.

Assessment method

The test and factors above, with a form that records the reasoning.

Notification rules

Which regulators and contracts apply, the timeframes, and who approves a notification.

Templates

Draft wording for individuals, regulators, customers, and staff, with gaps for the facts. Notices prepared in advance save days.

A breach register

Every breach, including those assessed as not notifiable. Patterns across several entries show where to invest.

Supplier terms

What contracts require suppliers to report, and how quickly.

Links to other plans

The incident response plan, the communications plan, and the insurance policy, with the point at which each is invoked.

Who Decides

Privacy or compliance lead

Leads the assessment and recommends the decision.

Incident lead and technical team

Contain the breach and establish what was accessed.

Forensic investigators

Provide the evidence of what was and was not taken.

Legal counsel

Advises on obligations, privilege, and contact with regulators.

Executive sponsor

Approves notifications and public statements.

Communications and customer service

Prepare notices and answer affected people.

In a small organization one person may hold several roles. The plan should still name who makes the notification decision, and who decides if that person is unavailable.

Obligations Beyond the Privacy Act

A single breach can trigger several duties at once, on top of the Notifiable Data Breaches scheme. Map them before an incident.

RegimeWho it applies toTimeframe
APRA incident reportingBanks, insurers, and superannuation funds regulated by APRA.As soon as possible, and within 72 hours of becoming aware of a material incident.
Critical infrastructure reportingEntities with obligations under the Security of Critical Infrastructure Act.Within 12 hours of becoming aware of a critical incident with a significant impact, and within 72 hours for an incident with a relevant impact.
Ransomware payment reportingBusinesses with annual turnover above AUD 3 million and critical infrastructure entities, under the Cyber Security Act 2024.Within 72 hours of making a payment. In force since May 30, 2025.
GDPR and UK GDPROrganizations holding personal data of people in the European Union or United Kingdom.To the supervisory authority within 72 hours of becoming aware, where feasible.
United States lawsOrganizations holding data of US residents, by state and sector.Varies by state and sector.
ContractsAny organization with notification clauses in customer or supplier agreements.Often shorter than the legal periods.

Data Breach Response With ThreatDefence

ThreatDefence provides the evidence a breach assessment depends on. Our responders establish which records were accessed or taken and over what period, and document the findings so they can support a notification decision.

We work alongside legal counsel and the insurer, help prepare the facts for a statement to the regulator, and keep the environment under 24x7 SOC monitoring during and after the response.

Forensic collection in the ThreatDefence portal: artifacts collected from a host, evidence by source, collection progress, fleet hunt results and findings

Experiencing a breach?

Our incident response team answers 24x7, with or without a retainer.

Talk to Our Experts

Discuss data breach readiness, forensic investigation, and 24x7 monitoring for your organization.

    I'm an MSP, MSSP, or reseller

    By submitting this form, you accept the terms in our privacy policy.

    Frequently Asked Questions

    Should a data breach also be reported to police or the ASD?

    A notification to the OAIC does not report the crime. Where a breach results from an attack, theft, or fraud, report it to police, and report the cyber incident to the Australian Signals Directorate through its ReportCyber service. Both reports are voluntary for most organizations and separate from the privacy notification. Critical infrastructure entities, and organizations that make a ransomware payment, have mandatory reports with their own timeframes.

    What are the steps in responding to a data breach?

    The steps are to contain the breach, assess the information involved and the risk of harm, notify those who need to know, and review the cause. Containment and assessment usually run together, and notification can begin before the assessment is finished.

    What is an eligible data breach in Australia?

    A breach is eligible when personal information has been accessed or disclosed without authorization, or lost where that is likely, when serious harm to at least one person is likely, and when remedial action has not removed that risk. A payroll file deleted unopened by a trusted recipient usually fails the test. A mailbox of passport scans synchronized by an attacker meets it.

    How long do we have to notify a data breach in Australia?

    The notification itself has no fixed number of days. Once there are reasonable grounds to believe an eligible data breach has occurred, the statement goes to the OAIC and individuals are told as soon as practicable. The 30 days applies earlier, to assessing a suspected breach, and is not permission to delay. An organization that already knows the breach is eligible notifies without waiting for day 30.

    Does every data breach have to be reported?

    No. In Australia only eligible data breaches are notified to the OAIC and to individuals. A breach contained before serious harm became likely, such as an encrypted laptop that never reconnected, is recorded with the reasoning and goes no further. Contracts, APRA, and overseas privacy laws apply their own tests, so a breach that is not eligible under the Privacy Act can still need reporting elsewhere.

    Who is responsible for data breach notification?

    The organization that holds the information is responsible. Where a supplier holds data on the organization’s behalf, both may have obligations, and the contract should say who notifies. The decision to notify normally sits with the privacy lead and an executive, on legal advice.

    What is the difference between a data breach and a cyber incident?

    A cyber incident is an event that compromises systems or security. A data breach involves information being accessed, disclosed, or lost without authorization, whether through a cyber incident or another cause. Many cyber incidents involve no data breach, and some breaches involve no attack.

    Should we notify before the investigation is complete?

    Yes, where there are already reasonable grounds to believe the breach is eligible, or where people need to act quickly to protect themselves, for example by replacing an identity document or changing a password. Send what is established, say plainly what is still being investigated, and commit to an update. Legal counsel should review the wording before it goes.