Skip to main content
Compliance guideReviewed

Cyber insurance in Australia: what four proposal forms ask about security testing

Chubb's Cyber ERM Short MarketPlace, Standard and Extensive forms and DUAL's 08.26 cyber form, read on 15 November 2026 and compared topic by topic: which of them ask about penetration testing and scanning, in their own words, and which of your records relate to each question.

Not insurance advice: your broker advises on cover.

Four proposal forms compared: testing questions and the records behind them

Eleven topics, one per row, compared across the four forms as each read on 15 November 2026. A form cell quotes the form or gives a faithful short form of it, with its section and question number; "None, as read" means that version has no question on the topic, apart from any nearby question the cell names.

How to use it

Read across a row to the records that relate, what a web application test report can show for that question, if anything, and what it cannot, and a step to take before answering. Every row reads only the forms and your own records and sends no request to any system. Work from the form your broker gives you: if its name, version or date differs from a column heading, check every question on it again. The records column and the test report column describe evidence, not a verdict: how an answer is read is the insurer's decision, and your broker advises on cover. None of the table is insurance or legal advice.

11 rows

Four proposal forms compared: testing questions and the records behind them

TopicChubb Short MarketPlace form (published 11/2023)Chubb Standard form (published 06/2023)Chubb Extensive form (footer reads "Published 06/0623")DUAL form (08.26)Records that relateWhat a web application test report can show, and what it cannotBefore you answer
1. Penetration testing of external or externally facing systemsNone, as read. Apart from a field for your website, none of the form's questions mentions penetration testing, vulnerability scanning, websites or web applications.Section V, question 3, the first item under "Please indicate if any of the following apply": "External penetration testing conducted at least annually". A tick box, not Yes or No.Section V, question 5, Yes or No: "Do you conduct penetration testing at least annually to assess the security of important externally facing systems?"None, as read. No question mentions penetration testing or vulnerability scanning. Section 1, question 7 asks for your website address.Your list of systems customers or the public can reach, marked with the ones you count as important. For each test: the testing dates, the report date and version, the scope as tested and its exclusions, the findings with their ratings, and the retest date, build and outcome for each finding.Can show: that the applications and hostnames in its scope were tested on its testing dates, what was found, and, after a retest, which findings were recorded as fixed on the retested build. Cannot show: any system outside its scope, that testing happens at least annually (one report records one test), or which systems the insurer reads as "external" or "important".Reads the forms and your own records only, and sends no request to any system. Mark each externally facing system as tested, with its testing dates, or not tested. Neither the Standard nor the Extensive question asks who performed the test, and neither of those forms' glossaries defines penetration testing. Take any doubt about which systems the words reach to your broker before you tick or answer.
2. Penetration testing of internal systemsNone, as read.Section V, question 3, the second item to tick: "Internal system penetration testing conducted at least annually".Section V, question 6, Yes or No: "Do you conduct penetration testing on important internal systems at least annually?"None, as read.The report from an internal test, from the provider that ran it, with its dates, its scope and any retest outcomes.Cannot show: internal networks, servers, staff devices or directory services, which sit outside a web application test. Whether an internet-facing application that only staff sign in to counts as an internal system for this question is one to ask your broker.Reads the forms and your own records only, and sends no request to any system. Answer from an internal test report, or record that none exists. Clever Ops does not test internal networks; that work goes to a network specialist.
3. Web application firewall in front of external applicationsNone, as read.Section V, question 3, the third item to tick: "Web Application Firewalls (WAF) are applied in front of most critically external facing applications".Section V, question 7, Yes or No: "Do you have a Web Application Firewall (WAF) in front of critical externally facing applications?"None, as read.A dated statement or export from your host, content delivery network or firewall provider listing which hostnames sit behind the firewall. The Standard and Extensive glossaries define a WAF as a firewall "placed between an application and the Internet".Cannot show: which applications sit behind a firewall. That is a configuration fact your provider records, not a test result.Reads the forms and your own records only, and sends no request to any system. Ask your provider or developer for the list rather than opening a settings screen that may also show keys. Note any externally facing application that is not behind the firewall.
4. Vulnerability scanning: coverage, frequency and toolsNone, as read.Section V, question 27(b): "What percentage of your assets are within scope for vulnerability scanning?" Question 28: "How often do you perform vulnerability scans?", with one space for Internal and one for External.Section V, question 48(b), the same percentage question. Questions 51 and 52: how often you perform internal, and external, vulnerability scans. Question 53: which vulnerability management tools you use for external and for internal scanning.None, as read.The asset inventory the percentage is worked out from, the scanning schedule that is actually running, and the dated scan results.Cannot show: how often scans run, or what share of your assets they cover. A penetration test report records one test on its dates.Reads the forms and your own records only, and sends no request to any system. Give the frequency the scan records show rather than a planned one, and work the percentage out from the inventory rather than estimating it.
5. Open portsNone, as read.Section V, question 31, Yes or No: "Do you regularly scan for and disable any unnecessary open ports and protocols?"Section V, question 55, Yes or No: "Do you regularly scan your external firewalls for any unnecessary open ports?" Question 56, Yes or No: "Do you disable all non-essential open ports and protocols?"None, as read.Your IT provider's or host's dated records of port checks and firewall changes.Cannot show: the ports open across your network. A web application test covers the hostnames and applications in its scope.Reads the forms and your own records only, and sends no request to any system. These questions are about your network as a whole, so the answer comes from whoever runs it.
6. Patch and remediation timelinesNone, as read.Section V, question 32: a formal patch management process, Yes or No. Question 33: target timelines by vulnerability criticality (CVSS), in days, for Low, Medium, High and Critical. Question 34: your level of compliance with these targets over the most recent 12 months. Question 35: what you do if a patch can not be applied in a timely manner.Section V, questions 57 to 60, the same four questions. Question 61, Yes or No: "Are patches tested in a controlled environment before deploying more broadly?"Section 10, question 1: describe your patch management process, including how you ensure that all critical patches are applied promptly, and the typical timeframe for deploying patches addressing zero day vulnerabilities once the vendor releases them. Question 2, Yes or No: "Are patches tested in a UAT environment before being deployed in the live environment?"Your written targets in days, and your update and release records. For findings from a test: each finding's CVSS score, the report date, the date its fix shipped, and the retest date and outcome.Can show: for the findings in one report, how long each stayed open between the report and a retest that recorded it as fixed, set against your own targets. Cannot show: updates to software and systems outside its scope, or findings nobody has retested.Reads the forms and your own records only, and sends no request to any system. Count a finding as closed only when a retest recorded it as fixed; partially fixed, still reproducible and not retested findings stay open on your records. The Standard and Extensive glossaries define CVSS without naming a version, so record which version your own targets and figures use.
7. Secure development and testing before releaseNone, as read.Section V, question 22, Yes, No or N/A: "Do your software developers receive training on the principles of writing secure applications?" Question 23: describe the quality control and testing procedures that apply to any new software programmes, including updates and new releases, on your network, including the minimal timeframe for a new or updated system to pass quality assurance testing before it is made operational on your live network, along with separate development, testing and acceptance environments.Section V, questions 43 (Yes or No) and 44, the same two questions.None on development, as read. Section 10, question 2 asks about testing patches in a UAT environment (row 6).Your release process as written and as followed, the environments a change passes through, and your developers' training records, with dates.Can show: which build and which environment the test ran against, such as a staging copy. Cannot show: that every release passes quality assurance, or that developers were trained.Reads the forms and your own records only, and sends no request to any system. Ask your developer or agency for the written process and the training records rather than describing them from memory.
8. Penetration testing of vendors, and vendor or supplier assessmentNone, as read. Section III, an optional Social Engineering Fraud extension, question 2, Yes or No: "Do you confirm all changes to vendor/supplier details (including routing numbers, account numbers, telephone numbers, and contact information) by a direct call using only the contract number previously provided by the vendor/supplier before the request was received?"Section VI, question 2: what is included in vendor assessments, "either prior to contracting or during audits". "Penetration testing" is one of the options to select.Section VI, question 6, the same question with the same option.Section 12, question 2, Yes, No or Partial: "Does the organisation conduct formal due diligence, including IT and cybersecurity due diligence, on all third party suppliers as part of its supplier risk management process?"For each vendor: what you asked it for and what it gave you, such as its own test summary, certifications or questionnaire answers, with dates.Cannot show: how a vendor secures or assesses its own product or platform. Your own test report covers your application, and in a Clever Ops test, services run by other businesses, such as a vendor's platform, stay out of scope unless their owner agrees.Reads the forms and your own records only, and sends no request to any system. Select only the steps your vendor files show were taken.
9. Network segmentation and its validationNone, as read, on segmenting your own network. Section II, question 5 asks a subsidiary, franchisee or smaller entity of a larger or parent organisation for details including any network interconnectivity and/or segmentation with the parent organisation.Section V, question 2: detail how your network has been structured or segmented to minimise lateral movement of malware or users, or the chance that multiple services are impacted by the same issue or vulnerability, and which methods it uses.Section V, question 3 (the network) and question 4 (how applications and systems are segregated), each with methods to select.Section 6, question 4, on "structured and documented network segmentation to clearly separate environments and restrict unauthorised lateral movement". Its answers are "Yes (segmented with enforced controls and regular validation)", "Partial (limited segmentation)" and "No".Network diagrams, and your IT provider's dated records of how segmentation is checked.Cannot show: how your network is segmented. A web application test does not validate segmentation inside a network.Reads the forms and your own records only, and sends no request to any system. The first DUAL answer mentions regular validation, so check that records of validation exist before choosing it.
10. "Tested" for plans and restores, not systemsNone, as read.Section IV, question 1(e): an incident response plan "that is tested at least annually". Section V, questions 13 and 14: business continuity and disaster recovery plans that address cyber scenarios, "tested annually". Question 16 lists backup testing practices, including full restore tests.Section IV, questions 1(f) and 1(g): incident response and data breach response plans tested at least annually. Section V, questions 29(a) and 30(a): "Is this tested at least annually?" for the continuity and recovery plans. Question 33: "Are full restore from backup tests performed at least annually?"Section 6, question 3: whether a Disaster Recovery Plan, a Business Continuity Plan and a Cyber Incident Response Plan are in place, and "Tested annually?" for each. Question 2: how often you validate restoration by performing a full restore using a sample set of back up data.The date and record of each plan exercise and each restore test.Cannot show: that a plan was exercised or a restore worked. A penetration test is not a test of these plans.Reads the forms and your own records only, and sends no request to any system. Answer each "tested" question from the exercise or restore record for that plan.
11. Incidents, breaches and claimsSection II, question 2, Yes or No: within the last 3 years, any Cyber Incidents, Data Breaches or Privacy Complaints, or awareness of any matter that could lead to a claim under a Cyber insurance Policy.Section VIII, question 1: listed events in the past five years, including Data Breach and Malicious Cyber Incident against you, leaving out events that existing security measures mitigated. Question 2: awareness of notices, facts, circumstances or situations that could qualify as a Data Breach, Cyber Incident or System Failure Event, among others.Section VIII, questions 1 and 2, the same. Section IV, question 2 also lists "Australia - Notifiable Data Breach Scheme" in a table with columns headed "Compliance Assessed in the past 12 months?", "Compliance Requirements Addressed?" and "Not Applicable".Section 13, questions 1 to 4: investigations, regulator requests, consent orders and complaints about personally identifiable information. Section 14, questions 1 and 2: claims, losses and circumstances that may give rise to a claim.Your incident register, breach assessments and any notifications made, with dates.Cannot show: incident history. A finding records a weakness the test found on the build tested, and whether anything in a report is a matter to tell the insurer about is a question for your broker.Reads the forms and your own records only, and sends no request to any system. Use each question's own period: 3 years on the Short form, the past five years in question 1 of the Standard and Extensive loss history sections, and the wording of each DUAL question. The Chubb Standard and Extensive glossaries define Cyber Incident and Data Breach.

Topic

Chubb Short MarketPlace form (published 11/2023)

  1. 1. Penetration testing of external or externally facing systems
    Chubb Short MarketPlace form (published 11/2023)
    None, as read. Apart from a field for your website, none of the form's questions mentions penetration testing, vulnerability scanning, websites or web applications.
    Chubb Standard form (published 06/2023)
    Section V, question 3, the first item under "Please indicate if any of the following apply": "External penetration testing conducted at least annually". A tick box, not Yes or No.
    Chubb Extensive form (footer reads "Published 06/0623")
    Section V, question 5, Yes or No: "Do you conduct penetration testing at least annually to assess the security of important externally facing systems?"
    DUAL form (08.26)
    None, as read. No question mentions penetration testing or vulnerability scanning. Section 1, question 7 asks for your website address.
    Records that relate
    Your list of systems customers or the public can reach, marked with the ones you count as important. For each test: the testing dates, the report date and version, the scope as tested and its exclusions, the findings with their ratings, and the retest date, build and outcome for each finding.
    What a web application test report can show, and what it cannot
    Can show: that the applications and hostnames in its scope were tested on its testing dates, what was found, and, after a retest, which findings were recorded as fixed on the retested build. Cannot show: any system outside its scope, that testing happens at least annually (one report records one test), or which systems the insurer reads as "external" or "important".
    Before you answer
    Reads the forms and your own records only, and sends no request to any system. Mark each externally facing system as tested, with its testing dates, or not tested. Neither the Standard nor the Extensive question asks who performed the test, and neither of those forms' glossaries defines penetration testing. Take any doubt about which systems the words reach to your broker before you tick or answer.
  2. 2. Penetration testing of internal systems
    Chubb Short MarketPlace form (published 11/2023)
    None, as read.
    Chubb Standard form (published 06/2023)
    Section V, question 3, the second item to tick: "Internal system penetration testing conducted at least annually".
    Chubb Extensive form (footer reads "Published 06/0623")
    Section V, question 6, Yes or No: "Do you conduct penetration testing on important internal systems at least annually?"
    DUAL form (08.26)
    None, as read.
    Records that relate
    The report from an internal test, from the provider that ran it, with its dates, its scope and any retest outcomes.
    What a web application test report can show, and what it cannot
    Cannot show: internal networks, servers, staff devices or directory services, which sit outside a web application test. Whether an internet-facing application that only staff sign in to counts as an internal system for this question is one to ask your broker.
    Before you answer
    Reads the forms and your own records only, and sends no request to any system. Answer from an internal test report, or record that none exists. Clever Ops does not test internal networks; that work goes to a network specialist.
  3. 3. Web application firewall in front of external applications
    Chubb Short MarketPlace form (published 11/2023)
    None, as read.
    Chubb Standard form (published 06/2023)
    Section V, question 3, the third item to tick: "Web Application Firewalls (WAF) are applied in front of most critically external facing applications".
    Chubb Extensive form (footer reads "Published 06/0623")
    Section V, question 7, Yes or No: "Do you have a Web Application Firewall (WAF) in front of critical externally facing applications?"
    DUAL form (08.26)
    None, as read.
    Records that relate
    A dated statement or export from your host, content delivery network or firewall provider listing which hostnames sit behind the firewall. The Standard and Extensive glossaries define a WAF as a firewall "placed between an application and the Internet".
    What a web application test report can show, and what it cannot
    Cannot show: which applications sit behind a firewall. That is a configuration fact your provider records, not a test result.
    Before you answer
    Reads the forms and your own records only, and sends no request to any system. Ask your provider or developer for the list rather than opening a settings screen that may also show keys. Note any externally facing application that is not behind the firewall.
  4. 4. Vulnerability scanning: coverage, frequency and tools
    Chubb Short MarketPlace form (published 11/2023)
    None, as read.
    Chubb Standard form (published 06/2023)
    Section V, question 27(b): "What percentage of your assets are within scope for vulnerability scanning?" Question 28: "How often do you perform vulnerability scans?", with one space for Internal and one for External.
    Chubb Extensive form (footer reads "Published 06/0623")
    Section V, question 48(b), the same percentage question. Questions 51 and 52: how often you perform internal, and external, vulnerability scans. Question 53: which vulnerability management tools you use for external and for internal scanning.
    DUAL form (08.26)
    None, as read.
    Records that relate
    The asset inventory the percentage is worked out from, the scanning schedule that is actually running, and the dated scan results.
    What a web application test report can show, and what it cannot
    Cannot show: how often scans run, or what share of your assets they cover. A penetration test report records one test on its dates.
    Before you answer
    Reads the forms and your own records only, and sends no request to any system. Give the frequency the scan records show rather than a planned one, and work the percentage out from the inventory rather than estimating it.
  5. 5. Open ports
    Chubb Short MarketPlace form (published 11/2023)
    None, as read.
    Chubb Standard form (published 06/2023)
    Section V, question 31, Yes or No: "Do you regularly scan for and disable any unnecessary open ports and protocols?"
    Chubb Extensive form (footer reads "Published 06/0623")
    Section V, question 55, Yes or No: "Do you regularly scan your external firewalls for any unnecessary open ports?" Question 56, Yes or No: "Do you disable all non-essential open ports and protocols?"
    DUAL form (08.26)
    None, as read.
    Records that relate
    Your IT provider's or host's dated records of port checks and firewall changes.
    What a web application test report can show, and what it cannot
    Cannot show: the ports open across your network. A web application test covers the hostnames and applications in its scope.
    Before you answer
    Reads the forms and your own records only, and sends no request to any system. These questions are about your network as a whole, so the answer comes from whoever runs it.
  6. 6. Patch and remediation timelines
    Chubb Short MarketPlace form (published 11/2023)
    None, as read.
    Chubb Standard form (published 06/2023)
    Section V, question 32: a formal patch management process, Yes or No. Question 33: target timelines by vulnerability criticality (CVSS), in days, for Low, Medium, High and Critical. Question 34: your level of compliance with these targets over the most recent 12 months. Question 35: what you do if a patch can not be applied in a timely manner.
    Chubb Extensive form (footer reads "Published 06/0623")
    Section V, questions 57 to 60, the same four questions. Question 61, Yes or No: "Are patches tested in a controlled environment before deploying more broadly?"
    DUAL form (08.26)
    Section 10, question 1: describe your patch management process, including how you ensure that all critical patches are applied promptly, and the typical timeframe for deploying patches addressing zero day vulnerabilities once the vendor releases them. Question 2, Yes or No: "Are patches tested in a UAT environment before being deployed in the live environment?"
    Records that relate
    Your written targets in days, and your update and release records. For findings from a test: each finding's CVSS score, the report date, the date its fix shipped, and the retest date and outcome.
    What a web application test report can show, and what it cannot
    Can show: for the findings in one report, how long each stayed open between the report and a retest that recorded it as fixed, set against your own targets. Cannot show: updates to software and systems outside its scope, or findings nobody has retested.
    Before you answer
    Reads the forms and your own records only, and sends no request to any system. Count a finding as closed only when a retest recorded it as fixed; partially fixed, still reproducible and not retested findings stay open on your records. The Standard and Extensive glossaries define CVSS without naming a version, so record which version your own targets and figures use.
  7. 7. Secure development and testing before release
    Chubb Short MarketPlace form (published 11/2023)
    None, as read.
    Chubb Standard form (published 06/2023)
    Section V, question 22, Yes, No or N/A: "Do your software developers receive training on the principles of writing secure applications?" Question 23: describe the quality control and testing procedures that apply to any new software programmes, including updates and new releases, on your network, including the minimal timeframe for a new or updated system to pass quality assurance testing before it is made operational on your live network, along with separate development, testing and acceptance environments.
    Chubb Extensive form (footer reads "Published 06/0623")
    Section V, questions 43 (Yes or No) and 44, the same two questions.
    DUAL form (08.26)
    None on development, as read. Section 10, question 2 asks about testing patches in a UAT environment (row 6).
    Records that relate
    Your release process as written and as followed, the environments a change passes through, and your developers' training records, with dates.
    What a web application test report can show, and what it cannot
    Can show: which build and which environment the test ran against, such as a staging copy. Cannot show: that every release passes quality assurance, or that developers were trained.
    Before you answer
    Reads the forms and your own records only, and sends no request to any system. Ask your developer or agency for the written process and the training records rather than describing them from memory.
  8. 8. Penetration testing of vendors, and vendor or supplier assessment
    Chubb Short MarketPlace form (published 11/2023)
    None, as read. Section III, an optional Social Engineering Fraud extension, question 2, Yes or No: "Do you confirm all changes to vendor/supplier details (including routing numbers, account numbers, telephone numbers, and contact information) by a direct call using only the contract number previously provided by the vendor/supplier before the request was received?"
    Chubb Standard form (published 06/2023)
    Section VI, question 2: what is included in vendor assessments, "either prior to contracting or during audits". "Penetration testing" is one of the options to select.
    Chubb Extensive form (footer reads "Published 06/0623")
    Section VI, question 6, the same question with the same option.
    DUAL form (08.26)
    Section 12, question 2, Yes, No or Partial: "Does the organisation conduct formal due diligence, including IT and cybersecurity due diligence, on all third party suppliers as part of its supplier risk management process?"
    Records that relate
    For each vendor: what you asked it for and what it gave you, such as its own test summary, certifications or questionnaire answers, with dates.
    What a web application test report can show, and what it cannot
    Cannot show: how a vendor secures or assesses its own product or platform. Your own test report covers your application, and in a Clever Ops test, services run by other businesses, such as a vendor's platform, stay out of scope unless their owner agrees.
    Before you answer
    Reads the forms and your own records only, and sends no request to any system. Select only the steps your vendor files show were taken.
  9. 9. Network segmentation and its validation
    Chubb Short MarketPlace form (published 11/2023)
    None, as read, on segmenting your own network. Section II, question 5 asks a subsidiary, franchisee or smaller entity of a larger or parent organisation for details including any network interconnectivity and/or segmentation with the parent organisation.
    Chubb Standard form (published 06/2023)
    Section V, question 2: detail how your network has been structured or segmented to minimise lateral movement of malware or users, or the chance that multiple services are impacted by the same issue or vulnerability, and which methods it uses.
    Chubb Extensive form (footer reads "Published 06/0623")
    Section V, question 3 (the network) and question 4 (how applications and systems are segregated), each with methods to select.
    DUAL form (08.26)
    Section 6, question 4, on "structured and documented network segmentation to clearly separate environments and restrict unauthorised lateral movement". Its answers are "Yes (segmented with enforced controls and regular validation)", "Partial (limited segmentation)" and "No".
    Records that relate
    Network diagrams, and your IT provider's dated records of how segmentation is checked.
    What a web application test report can show, and what it cannot
    Cannot show: how your network is segmented. A web application test does not validate segmentation inside a network.
    Before you answer
    Reads the forms and your own records only, and sends no request to any system. The first DUAL answer mentions regular validation, so check that records of validation exist before choosing it.
  10. 10. "Tested" for plans and restores, not systems
    Chubb Short MarketPlace form (published 11/2023)
    None, as read.
    Chubb Standard form (published 06/2023)
    Section IV, question 1(e): an incident response plan "that is tested at least annually". Section V, questions 13 and 14: business continuity and disaster recovery plans that address cyber scenarios, "tested annually". Question 16 lists backup testing practices, including full restore tests.
    Chubb Extensive form (footer reads "Published 06/0623")
    Section IV, questions 1(f) and 1(g): incident response and data breach response plans tested at least annually. Section V, questions 29(a) and 30(a): "Is this tested at least annually?" for the continuity and recovery plans. Question 33: "Are full restore from backup tests performed at least annually?"
    DUAL form (08.26)
    Section 6, question 3: whether a Disaster Recovery Plan, a Business Continuity Plan and a Cyber Incident Response Plan are in place, and "Tested annually?" for each. Question 2: how often you validate restoration by performing a full restore using a sample set of back up data.
    Records that relate
    The date and record of each plan exercise and each restore test.
    What a web application test report can show, and what it cannot
    Cannot show: that a plan was exercised or a restore worked. A penetration test is not a test of these plans.
    Before you answer
    Reads the forms and your own records only, and sends no request to any system. Answer each "tested" question from the exercise or restore record for that plan.
  11. 11. Incidents, breaches and claims
    Chubb Short MarketPlace form (published 11/2023)
    Section II, question 2, Yes or No: within the last 3 years, any Cyber Incidents, Data Breaches or Privacy Complaints, or awareness of any matter that could lead to a claim under a Cyber insurance Policy.
    Chubb Standard form (published 06/2023)
    Section VIII, question 1: listed events in the past five years, including Data Breach and Malicious Cyber Incident against you, leaving out events that existing security measures mitigated. Question 2: awareness of notices, facts, circumstances or situations that could qualify as a Data Breach, Cyber Incident or System Failure Event, among others.
    Chubb Extensive form (footer reads "Published 06/0623")
    Section VIII, questions 1 and 2, the same. Section IV, question 2 also lists "Australia - Notifiable Data Breach Scheme" in a table with columns headed "Compliance Assessed in the past 12 months?", "Compliance Requirements Addressed?" and "Not Applicable".
    DUAL form (08.26)
    Section 13, questions 1 to 4: investigations, regulator requests, consent orders and complaints about personally identifiable information. Section 14, questions 1 and 2: claims, losses and circumstances that may give rise to a claim.
    Records that relate
    Your incident register, breach assessments and any notifications made, with dates.
    What a web application test report can show, and what it cannot
    Cannot show: incident history. A finding records a weakness the test found on the build tested, and whether anything in a report is a matter to tell the insurer about is a question for your broker.
    Before you answer
    Reads the forms and your own records only, and sends no request to any system. Use each question's own period: 3 years on the Short form, the past five years in question 1 of the Standard and Extensive loss history sections, and the wording of each DUAL question. The Chubb Standard and Extensive glossaries define Cyber Incident and Data Breach.

Cyber insurance in Australia and the insurer's proposal form

The four cyber insurance forms this guide reads are each titled a proposal form.

Under "Completing This Proposal Form", each of the three Chubb forms read here says "This document allows Chubb to gather the needed information to assess the risks related to your information systems." DUAL's asks the applicant to answer all questions giving full and complete answers, and to contact their broker if they have any doubt about what a relevant fact is. Whatever goes on the form is your business's own written account of itself.

This guide sets four of those forms side by side: Chubb Insurance Australia's Cyber ERM Short MarketPlace, Standard and Extensive forms, and DUAL Australia's cyber proposal form marked 08.26. It shows which of them ask about penetration testing and vulnerability scanning, in the forms' own words and question numbers, and which of your records relate to each question: the dates of a test, its scope, the fixes that followed and the retest.

The comparison table on this page is the working part, and the sections explain how to read it.

This guide gives neither insurance advice nor legal advice. What cover suits your business, what your policy says and how an insurer weighs any answer are matters for your broker and your insurer. If you came here looking for cyber insurance requirements, the nearest thing this guide can offer is what these forms ask, in their own words.

Two of the four ask about penetration testing and two do not.

That is a reading of four documents from Chubb and DUAL as they stood on 15 November 2026, not a statement about cyber insurance in general, and it says nothing about what any policy requires.

The four forms, and the versions read on 15 November 2026

Each form was downloaded from its publisher's own website, Chubb's or DUAL's, on 15 November 2026 and read in full.

A proposal form can be revised, so every statement on this page is about the version named below as it read that day. The form your broker sends you is the one to answer: check its name, its version or date and its revenue line against this list before you rely on the comparison.

  • Chubb Cyber ERM Short MarketPlace Cyber Proposal Form, Australia. Its footer reads "Published 11/2023". Its stated band is revenue of less than $50m, and its underwriting questions 2 to 10 take a Yes or No answer.
  • Chubb Cyber ERM Standard Cyber Proposal Form, Australia. Its footer reads "Published 06/2023", and it says it is for businesses with revenue between $50m and $700m. Its Section V is headed "Technical Controls and Processes".
  • Chubb Cyber ERM Extensive Cyber Proposal Form, Australia. Its footer reads "Published 06/0623", exactly as printed, and the same page carries a 2023 copyright line. It says it is for businesses with revenue of more than $700m, and its Section V carries the same heading.
  • DUAL Australia Proposal Form: Cyber Insurance 08.26, the words at the foot of its numbered pages. Its cover reads "Revenues between $100m and $1bn", and its questions run through sections numbered 1 to 16, from details of the applicant to operational technology.

Start with the revenue line

Each form states the revenue band it is written for, and the four bands differ. For cyber insurance for a smaller business, the Short MarketPlace form is the Chubb form read here that is written for revenue of less than $50m. Ask your broker which form your business will be given: a form for another band, or a later version of one of these, can ask different questions.

Which forms ask about penetration testing, and in what words

As read on 15 November 2026, the Chubb Standard and Extensive forms ask about penetration testing, and the Short MarketPlace and DUAL forms do not.

On the Standard form it is a tick list. Section V, question 3 says "Please indicate if any of the following apply" and lists these items, each with a box to tick: "External penetration testing conducted at least annually", "Internal system penetration testing conducted at least annually" and "Web Application Firewalls (WAF) are applied in front of most critically external facing applications".

The Extensive form covers the same ground with separate Yes or No questions in Section V. Question 5 asks "Do you conduct penetration testing at least annually to assess the security of important externally facing systems?", question 6 asks "Do you conduct penetration testing on important internal systems at least annually?", and question 7 asks "Do you have a Web Application Firewall (WAF) in front of critical externally facing applications?"

The wording differs in ways that change what an accurate answer needs. The Standard form's item reads "External penetration testing", while the Extensive form's question 5 is about "important externally facing systems", which leaves you to say which systems are important.

Both set the period as "at least annually", and both keep internal and external testing apart. Neither the Standard form's testing items nor the Extensive form's questions 5 and 6 ask who performed the testing or ask for a report, and neither form's Glossary of Defined Terms has an entry for penetration testing, although both define a web application firewall and the Common Vulnerability Scoring System (CVSS).

In the versions read that day, the Short MarketPlace form, apart from a field for your website, has no question that mentions penetration testing, vulnerability scanning, websites or web applications. The DUAL form has no question that mentions penetration testing or vulnerability scanning, and asks for your website address in Section 1, question 7.

Its closest questions are about patching in Section 10 and network segmentation in Section 6, question 4, both set out in the comparison.

Where "tested" means a plan or a restore

Three of the four forms read here use "tested" for something other than a system. The Standard form (Section IV, question 1(e), and Section V, questions 13 and 14), the Extensive form (Section IV, questions 1(f) and 1(g), and Section V, questions 29(a) and 30(a)) and the DUAL form (Section 6, question 3) ask whether incident response, data breach response, business continuity or disaster recovery plans are tested annually or at least annually.

The Extensive form's Section V, question 33 and the DUAL form's Section 6, question 2 ask about full restores from backup.

A penetration test report is not the record for any of these. Each is answered from the record of the exercise or the restore itself.

Before you tick: what the testing questions leave to your records

A tick on the Standard form and a Yes on the Extensive form each state something about your business.

Three points in the wording decide which records you need before giving either.

  • Which systems. "External" and "important externally facing systems" have no glossary entry on either form as read. Start from a list of every system customers or the public can reach, mark the ones you count as important and why, and mark which of them a test has covered. Both forms also ask you to describe the systems your business depends on most, internally and externally hosted (Standard, Section V, question 17; Extensive, Section V, question 38), so any externally facing system named in that answer belongs on this list too.
  • Which period. "At least annually" is about testing over time. The testing dates of each test you rely on show how far apart the tests were. If you have had one test, your records hold one set of dates, and that is worth raising with your broker before a tick is left to stand for a pattern of testing.
  • Which side of the network. Both forms ask about external and internal testing separately. A test of a web application that customers reach from the internet covers that application as the internet reaches it. Neither form defines "internal", so ask your broker before counting such a test towards the internal question.

Who reads the answer

Where the records leave an answer open, such as a system you are unsure counts as important, or a test whose dates sit close to the edge of the period, put the question to your broker with the records attached and keep the reply. How the insurer reads its own words is for the insurer to say.

Report dates, scope, remediation and retest: which record answers which part

Four kinds of record relate to the testing questions on the Standard and Extensive forms.

Each answers a different part of a question, and none answers all of it. The cyber insurance security checklist sets out how to read a test report as renewal evidence; this section ties each kind of record to the form questions it bears on.

Report and testing dates

The testing dates show when the work ran, and the report date and version identify the document you would produce if asked for it. For "at least annually", keep the dates of every test, not only the most recent one, since a single test's dates cannot show how often testing happens.

A report dated after the form is signed is not a record behind the answer given on the day of signing.

Scope as tested

The scope as tested lists the targets, the roles, the build and the dates, along with anything excluded or not performed. A Clever Ops report sets it out that way, and a report from any provider can be read for the same items. Set it against your list of externally facing systems: the systems it names are the ones the report speaks for.

A marketing site on a website builder, a booking page a vendor runs on one of your subdomains, or an API released after testing finished sits outside it unless the scope names it.

Remediation against your own targets

The Standard form asks for target timelines in days for Low, Medium, High and Critical vulnerabilities, by CVSS, and for your level of compliance with those targets over the most recent 12 months (Section V, questions 33 and 34). The Extensive form asks the same in Section V, questions 58 and 59.

The CVSS v3.1 specification's qualitative scale maps scores to ratings: None 0.0, Low 0.1 to 3.9, Medium 4.0 to 6.9, High 7.0 to 8.9 and Critical 9.0 to 10.0.

Low, Medium, High and Critical appear on that scale and on both forms, so a report that gives each rated finding a CVSS v3.1 score lets you set each finding's dates against your own targets: when it was reported, when the fix shipped and when a retest confirmed it. The glossaries on both forms define CVSS without naming a version, so record which version your own targets and figures use.

Retest outcomes

A Clever Ops retest records a status against every finding: fixed, partially fixed, still reproducible or not retested. Only the first closes a finding on your records. A compliance figure built from fixes that were reported but never confirmed would say more than the retest showed.

Answering from records: what each form's declaration says

Answering a proposal form is your business's own statement to its insurer, and each form's declaration sets out what the person signing declares.

The three Chubb declarations are worded alike. The authorised officers declare that "to the best of their knowledge and belief the statements made in this proposal and in all attachments and schedules to this proposal are true", and that "notice will be given as soon as practicable should any of the above information change between the date of this proposal and the proposed date of inception of the insurance".

Each Chubb declaration also says the proposal and its statements "shall be the basis of and will be incorporated in the policy should one be issued".

DUAL's declaration says the statement and particulars in the form are true and that no material facts have been misstated or suppressed after enquiry, that the undersigned will give immediate notice if any of the information alters between the date of the proposal and the inception date, and that the proposal, with any other information supplied, "shall form the basis of any contract of insurance effected thereon".

DUAL also recommends keeping a record of all information supplied for the purpose of entering into an insurance contract, including copies of the proposal form and correspondence.

Every form also prints a notice about the duty of disclosure. The Chubb Standard and Extensive forms use "You" for the Named Insured and all its Subsidiaries, and their notice says that before You enter into an insurance contract, You have a duty to tell Chubb anything that You know, or could reasonably be expected to know, may affect Chubb's decision to insure You and on what terms; that You have this duty until Chubb agrees to insure You; and that You have the same duty before You renew, extend, vary or reinstate an insurance contract. DUAL's notice, headed "Non-Disclosure", uses closely matching words.

Compilation No. 33 of the Insurance Contracts Act 1984 (Cth), on the Federal Register of Legislation, gives the text of the Act in force on 8 October 2026, its compilation date. In that text, the duty of disclosure is set out in section 21, within Division 1 of Part IV ("Disclosures, misrepresentations and use of certain information").

Under section 20E, Division 1 applies to contracts of insurance that are not consumer insurance contracts, and to proposed contracts that, if entered into, would not be consumer insurance contracts.

Consumer insurance contracts fall under Division 1A instead, where section 20B(1) says that, subject to the Act, "an insured has a duty to take reasonable care not to make a misrepresentation to the insurer before the relevant contract of insurance is entered into".

The cyber insurance security checklist gives these sections in more detail. Which duty applies to your contract, and what it means for your answers and your policy, is a question for your broker, and for a lawyer where it matters.

Wording that the records support

Answer each testing question from the records you hold on the day you sign, in words those records support, and never pick a wording because it reads better than the records do. A tick beside "External penetration testing conducted at least annually" when one application has had one test, or a Yes to the Extensive form's question 5 when one of your important externally facing systems has never been tested, states more than the records show.

Where the true position is partial, write it out in full and ask your broker how the insurer wants it given.

How to use the comparison table

The table compares the four forms topic by topic, with a column for each form, the records that relate, what a web application test report can show for that topic, if anything, and what it cannot, and a step to take before answering.

Every row reads only the forms and your own records, and none sends a request to any system. Work through it in this order.

  • Get the form your broker gives you and find its name, its version or published date and its revenue line. If they match a column heading, that column shows what the form asked on 15 November 2026. If they do not, treat the table as a list of topics to look for and read every question on your own form afresh.
  • For each row, copy the matching question from your form exactly as printed, with its section and question number.
  • Gather the records named in the records column, and write the date on each one.
  • Read the column on what a test report can show before relying on a report, and pass every question a report cannot answer to whoever holds the record: your IT provider, host, developer or finance team.
  • Write each answer in full before choosing a box. List what is still unsettled for your broker, send it before anyone signs, and file the reply with the signed form and the records each answer used.

Where a Clever Ops report relates to these questions, and where it does not

Clever Ops works on the parts of a business that customers reach through a browser or an API.

Its web application and API penetration test signs in to a custom application from test accounts in each role and tests the API behind it as well. The report sets out the scope as tested (targets, roles, build, dates and exclusions), gives each rated finding a CVSS v3.1 score alongside a business-risk rating, and orders the remediation plan by priority.

After the one consolidated retest, every finding carries a status. For an application inside that scope, those records relate to the external testing items on the Chubb Standard and Extensive forms and to their remediation questions.

The internal testing questions, the scanning questions and any system outside the scope are left for other records.

The website security check covers less: what anyone on the internet can see without signing in. Logged-in areas, source code and business logic sit outside it, which is why it is not a penetration test. If a form asks about penetration testing, a check report is not a record of one, so ask your broker whether the insurer wants to hear about the check at all.

Clever Ops does not test internal networks, servers, Active Directory or staff devices; that work belongs with a network specialist. An application Clever Ops built or hosts gets a security review rather than an independent test, and its report names who built it and who hosts it. Where a broker or insurer specifies a CREST-accredited tester, the work goes to a provider with that accreditation, because Clever Ops is not CREST accredited.

Answering the form stays with your business and your broker, how an answer is read is the insurer's decision, and Clever Ops does not provide incident response.

If your form asks about penetration testing and you decide to commission a test, bring that question to a free scoping call: its section and number, its exact words and the externally facing systems on your list. The written scope that follows names each target and each exclusion, so the finished report can be read against the systems your answer will cover.

Frequently asked questions

This guide cannot answer that for any policy, and two of the four forms it reads do not ask. As read on 15 November 2026, the Chubb Cyber ERM Standard and Extensive forms ask whether penetration testing is conducted at least annually, while the Short MarketPlace form and the DUAL 08.26 form have no question about it. What your own form asks is on the form your broker gives you, and what your policy says is for your broker to explain.

As read on 15 November 2026, the two Chubb forms that ask cover external and internal penetration testing at least annually, a web application firewall in front of the external facing applications each form describes, how often vulnerability scans are performed and what percentage of assets is within scope for them, open ports, and target timelines by vulnerability criticality (CVSS). The Short MarketPlace form asks none of these. The DUAL form has no question about penetration testing, vulnerability scanning, a web application firewall or open ports, but asks how patches are managed and whether they are tested in a UAT environment before being deployed in the live environment. The comparison table gives each question with its number.

Each form prints the band it is for. As read on 15 November 2026, the Short MarketPlace form is for businesses with revenue of less than $50m, the Standard form for revenue between $50m and $700m, and the Extensive form for revenue of more than $700m. Ask your broker which form to complete; the one you are given is the one to answer.

Not as a record of a penetration test. The Clever Ops website security check is limited to what is visible from the internet without signing in, so it holds no signed-in testing of an application. If a form asks about penetration testing, tell your broker what the check covered and ask whether the insurer wants to hear about it.

No. As read on 15 November 2026, neither the Standard form's tick boxes nor the Extensive form's questions 5 and 6 ask who performed the testing, and neither form's glossary defines penetration testing. If your broker or insurer asks for an independent tester, pick a provider with no part in building or hosting the application: where Clever Ops built or hosts it, the work is a security review and its report says so. If a CREST-accredited tester is specified, Clever Ops is not CREST accredited and refers the work to an accredited provider.

Then it is open on your records, and the answer comes from the retest record rather than from the plan to fix it. A finding recorded as partially fixed, still reproducible or not retested has not been confirmed fixed. If the position changes before cover starts, the Chubb declarations say notice will be given as soon as practicable, and DUAL's says immediate notice will be given, of information that changes between the date of the proposal and the inception date. Ask your broker how to present an open finding.

Scope the right test for your application

Tell us what you have built, who uses it and when it needs to launch. You will get a plain-English recommendation: a website audit, a web application test or a referral to a specialist.