You find out by putting two things side by side. On one side, the answers on your last proposal or renewal form. On the other, the configuration running this morning. Then you work down the list, writing an outcome, a date and a name against every line. A Cyber Strength Audit does that across seven areas of an environment.
The policy is filed. The match is not, and the question usually arrives from a board that has started asking what the renewal form actually claimed. Notice what is being asked. Not whether your controls are good, but whether the company's description of itself is still accurate. Usually nobody has checked.
Worth being plain about who does what. inSUPPORT does not sell or advise on cyber insurance and holds no Australian Financial Services Licence. Where cover forms part of a plan it is arranged through licensed insurance partners and underwritten by the insurer, subject to the insurer's assessment. So none of this is about your policy. It is about the IT and evidence job underneath it.
Three parts to the work: where the gap opens during a year of ordinary operating, what a comparison compares, and why the answer stops being true unless somebody keeps it true. That last part has an owner whether or not anyone has named one. An IT environment gap analysis gets filed as a paperwork job. It behaves like a maintenance one.
TL;DR: What you can actually check
- ✅ The form is a dated snapshot. The environment is not. A year of ordinary changes separates them.
- ✅ Two columns: what was declared and what is configured today, with an outcome and a date on every line.
- ✅ Borrow the arithmetic the Australian Signals Directorate uses. A line nobody has checked yet counts as not assessed, which is neutral and separate from not effective. Where several actions sit under one principle, the principle takes the lowest of their ratings.
- ✅ Cover is the insurer's decision. Whether the record still describes the environment is yours.
Contents
- Where the Gap Opens in Twelve Months
- What a Comparison Actually Compares
- Why the Answer Has a Short Shelf Life
- Straight answers before you start the comparison
Where the Gap Opens in Twelve Months
A proposal form is a photograph. It records what somebody believed was true about your environment on the day they filled it in, and then it goes into a folder and stops updating. The environment carries on without it.
Almost everything that moves the two apart is dull. A team buys a cloud tool on a company card and nobody tells IT. A site gets absorbed and its old kit comes along with it. A contractor is given administrator rights for a three-week project and the account is still live in December. A finance system comes out of the patch window after one update broke it, and the exception quietly becomes permanent because nothing bad happened. None of that is negligence. It is a year of a business operating, and each item is one more line where the declaration and the configuration have quietly stopped agreeing. That is the whole question.
Which is why the first useful move is not technical at all. Write down what changed in the last twelve months, before you look at a single control. New sites, new systems, new suppliers, and everybody who gained elevated access. That list takes an afternoon. It tells you where the comparison is going to hurt, which is the difference between a targeted check and an open-ended sweep of everything you own. Nobody outside the business can write it for you.
The Side You Control Moves
Closing that distance is not an unusual standard somebody invented for insurance purposes. It is what current Australian guidance already describes as the job. The Australian Signals Directorate's assessment evidence guide for enterprise IT, read on 16 November 2026, carries an action on exactly this. Keep checking that your suppliers still do what they promised on security, and do something about it when the proof runs out or the supplier's risk changes. ASD's wording is to "maintain current assurance that suppliers continue to meet their cyber security commitments, and act when that assurance lapses or the supplier's risk changes". Taking no action when it lapses is listed there as a sign the work is not effective.
The same guide treats old systems the same way. If there is no formal list of the ageing systems still running in your environment, ASD counts that as a sign the work is not effective. Its wording is "legacy systems are not formally identified within the environment". Both of those are drift with a different name on it. If nobody is watching for them, they surface at the next audit, or at the next time somebody asks you to prove something.
The Side You Do Not Control Moves Too
Your environment is only half of it. The questionnaire gets longer and more technical from one year to the next, and the language the questions are built on moves as well. On 15 October 2026 the Australian Signals Directorate published the ASD Essentials, a principles-based framework, alongside the Essential 8 that insurer questionnaires have been quoting for years.
Two things are worth stating precisely, and both were read at the source on 16 November 2026. ASD's published answer on timing is that it will advise on expectations for transitioning to the new framework in due course. That answer contains no date. ASD also states who the series is written for, primarily: Commonwealth entities, critical infrastructure owners and operators, and large organisations. In the same answer it says smaller businesses can still use the guidance where relevant, and that ASD publishes tailored guidance for them separately. ASD's own evidence guide, published the same day, still lists its Essential 8 guidance dated 2 October 2024 among the frameworks where supporting evidence may already exist.
So the instruction that follows is narrow and slightly boring. Answer the form in the words you can evidence today. Do not reach for newer framework vocabulary to describe work nobody has done yet. A declaration gets read later against what was actually in place, not against the sophistication of the wording.
What a Comparison Actually Compares
Two columns, and the discipline is in refusing to skip a row. The left column is the answers as they were submitted. The right column is what the environment shows today. Every row ends up with an outcome, the date it was checked and the name of whoever checked it. A row with no date on it is a memory.
The right-hand column is also the one the other side may reach into. On the Insurance Council of Australia's account, writing a cyber policy can require insurers to access a client's data and IT processes and potentially test their cyber defences. The reason it gives is that the ever-changing nature of the risk means coverage cannot be predicted from prior historical claims experience. Read that as a statement about what is being assessed. It is not only the form. It is the environment the form describes.
The paperwork half needs no licence and no technical background. It needs somebody willing to treat each answer as a factual question with a current answer, and to write down the ones that come back wrong. The right-hand column is a different job. A configuration export, a tested restore and a current list of who holds administrator rights come from whoever runs the environment. A penetration test comes from someone qualified to run one. A spreadsheet holds the comparison. It does not produce the evidence.
Read the Form as Statements of Fact
Every tick on that form is a row waiting for an outcome. Multi-factor authentication, with the scope it was enforced at. Backups, with the date of the last restore somebody actually completed. A patch window, with the exceptions that have since been granted against it. Administrator rights, with the current list of who holds them. Each of those was an answer about configuration, and the person who gave it may have been describing the intention, or may not have been placed to tell the difference. That is worth knowing before you decide what a gap means about anybody. Blame is a separate job.
It also helps to know why those particular questions exist, which our piece on what an underwriter weighs before covering an Australian business works through. If your form is still unanswered rather than submitted last year, start there instead.
Seven Places the Audit Looks
What goes into the right-hand column is published rather than vague, which makes it checkable. A Cyber Strength Audit examines five areas of the environment directly: identity and access, patching and currency, backup and recovery, email and endpoint, and configuration drift. It runs internal and external penetration testing. And it covers insurance alignment, which in inSUPPORT's own published words means that where a policy is already held, it is read against what is technically in place. That is the seventh area.
That last item is the one doing the work here, because it is the only one that points back at the left-hand column. The other six establish what is true today. A stack of technical findings is not yet an answer to the board's question. Reading the policy against what is configured is what makes it one.
Write Down What Nobody Could See
Borrow the arithmetic an assessor uses, because it stops the comparison being graded on optimism. In ASD's published answers, read on 16 November 2026, each action is rated not effective, partially effective or fully effective. An action nobody has assessed yet is rated not assessed. That is a neutral result, and a different thing from not effective. Where several actions sit under one principle, the principle takes the lowest of their ratings rather than the average. Applied to your two columns, a row nobody could verify becomes an outcome in its own right, with an owner and a date against it. One weak control sets the result for everything resting on it.
The same guide is explicit that effectiveness has to be considered in the context of each individual domain rather than across the organisation. A set of controls that works for one does not automatically guarantee effectiveness for another. In plain terms, keep the scope attached to every row. Enforced on the head office and not on the two sites you bought last year is a different answer from enforced. The form did not have a box for that.
Accuracy is the half of this you control. The reason it matters is statutory rather than tidiness. For a commercial policy, the Insurance Contracts Act 1984 (Cth) places a duty on the insured to disclose, before the contract of insurance is entered into, every matter of the kinds the Act describes that is relevant to the insurer's decision. It is a duty to tell the insurer what is relevant, which is wider than getting a form's answers right. An answer that has stopped describing the environment is one of the ways that duty gets tested, and it is the way you can check for yourself. Everything downstream of it belongs to somebody else.
Why the Answer Has a Short Shelf Life
A comparison answers the question for the week it was run. On the Monday somebody buys a tool, grants an access or defers an update, and the drift restarts from the new baseline. That is not an argument against running one. It is the reason a comparison on its own is worth less than the effort it takes. Left alone, each renewal starts further from the record than the last one, and the work of answering it honestly gets heavier rather than lighter every year.
So the decision in front of you is not really whether to check. It is who owns the match between checks, and whether that is the same party that runs the environment. If the honest answer is nobody, the next form gets filled in from memory as well, and the board's question about whether the cover would respond comes back next year wearing the same clothes.
The 6 Step Cyber Strength System puts that in two of its six steps. Comply produces documented evidence of the alignment rather than verbal assurance. Support keeps that alignment in place, with configuration drift monitored continuously rather than discovered at the next audit, and with the maintenance work carried on when policy wording changes. An IT environment gap analysis is a standing job on that model, not a week in the calendar.
The Audit Is the Baseline, Not the Answer
Treat the first comparison as the baseline it is. Expect it to find things, and expect some of them to be awkward. An environment nobody has measured against a declaration has had no reason to line up with one. A first pass that turns up nothing at all is the result worth questioning hardest.
Then date the document and keep it. The second version starts from the first one instead of from memory, which is the entire difference between a record and an annual panic. It is also what makes the next renewal a shorter conversation.
Who Owns the Record Between Audits
This is where the operating model stops being marketing and starts mattering. Where the IT, the compliance work and the insurance pathway run as one model, the evidence comes out of the work as a by-product. The party that changes a setting is the party that records it. Where those three sit with three different parties, somebody has to assemble the record from three directions at renewal. That assembly usually runs on other people's notes, in the fortnight before it is due.
You can test which one you are on without a technical conversation. Ask who owns the match between audits and ask for the answer in writing. Then ask what the cadence is, and attach it to the two dates that already exist in your year, the renewal and the board calendar. Both are already in the diary.
What a Finding Costs You to Close
There is a commercial question sitting underneath all of this and it is better asked out loud. Ask it before the comparison, not after. When the comparison finds a gap, who pays to close it? If the answer is that a quote arrives, then every mismatch anybody finds is a new project for whoever found it. The incentive runs in an awkward direction.
Plenty of providers work exactly that way and there is nothing improper about it. The only real mistake is discovering which model you are on after the gap list lands rather than before it. On the inSUPPORT model, the remediation an audit finds is included in the support fee rather than billed back as a separate project. Insurance, backup and security-awareness training are separate, clearly-quoted line items. You see which is which.
Straight answers before you start the comparison
How do we check whether our IT environment still matches what our cyber insurer was told?
An IT environment gap analysis works from the submitted form, not from memory. Get the proposal or renewal form exactly as it went back. Read each answer as a statement about configuration. Then verify each one against the live environment, and record the outcome, the date and the person on every line. Rows nobody can verify still get an outcome. Mark them unverified and put a name against each, because an unverified row you can see is governable and a blank one is not. The comparison is finished when every line carries something, not when every line is clean.
Who should own the match between one comparison and the next?
Ask, then read the answer carefully, because this is where a reassuring reply and a real one sound alike. A real answer names a party, says whether the work sits inside what you already pay or outside it, and survives being written down. A reassuring answer uses the word monitoring and stops there. Three replies are worth testing. If it is nobody, the next form gets filled in from memory. If it is you, ask yourself honestly who tells you when a setting changed. If it is the provider, ask them to show you how last month's changes reached the record, because a party that cannot show you March cannot have been keeping November true. Anything about the policy itself belongs with the licensed party that arranged the cover, and cover is subject to the insurer's assessment.
How long does it take, and what does a good result look like?
Your twelve-month change list sets the duration, which is why it is worth writing before you ask anyone for a quote. A business that added two sites, a finance system and four administrators has more rows to verify than one that changed nothing. The list tells you which you are. Treat a fixed quote from someone who has not seen that list, or your environment, with suspicion. On the result, do not judge it by how few findings it produced. Judge it by whether every declared control now carries an outcome, every gap carries an owner and a date, and the document says when it was true.
Where do we start if nobody has ever done this?
Start with the twelve-month change list, because it is the one input nobody outside the business can hand you and it is the fastest thing on the list. Every new site, system, supplier and person who gained elevated access since the last form went back, in one document. Then ask whoever holds the insurance paperwork for the form as it was actually submitted, so the list has something to be compared against. Those two together tell you which controls to verify first, and that ordering is what keeps the technical half from becoming a sweep of everything you own.
A Cyber Strength Audit is the instrument that builds the right-hand column. It examines the environment area by area, and where a policy is already held it is read against what is technically in place. What comes back is a plain-English risk picture, ranked by what would actually hurt. Then a documented audit trail you can hand to a board, an insurer or a client who asks, and a costed path to closing each gap. That ordering is deliberate, because a picture with no path is the thing this industry already sells plenty of. Cover stays the insurer's call. Everything on your side of that line can be settled.
Book a Cyber Strength Audit →Citations
- "Frequently asked questions, Australian Signals Directorate", ASD's own answers on the ASD Essentials series. The source for the statement that the series is written primarily for Commonwealth entities, critical infrastructure owners and operators, and large organisations, with tailored guidance published separately for small and medium businesses, and for ASD's position that it will advise on expectations for transitioning to the new framework in due course. Also the source for the effectiveness ratings used above: an action is rated not effective, partially effective or fully effective, not assessed is allocated to actions that have yet to be assessed, and the overall effectiveness rating for a principle is determined by the lowest applicable action rating. Published and last reviewed 15 October 2026. Read 16 November 2026. cyber.gov.au
- "Assessment evidence guide for enterprise IT, Australian Signals Directorate", The source for the statement that a set of controls effective in one domain does not automatically guarantee effectiveness in another, for the supplier assurance and legacy system indicators quoted above, and for the framework table that lists ASD's Essential 8 guidance dated 2 October 2024 among the places supporting evidence may already exist. Published and last reviewed 15 October 2026. Read 16 November 2026. cyber.gov.au
- "Insurance Contracts Act 1984 (Cth), Compilation No. 33", The source for the insured's duty of disclosure at section 21, which requires disclosure before the relevant contract of insurance is entered into of every matter known to the insured that is relevant to the insurer's decision whether to accept the risk and on what terms. Section 21 sits in Division 1, and section 20E applies that Division to contracts of insurance that are not consumer insurance contracts, which is why section 21 governs a commercial cyber policy. Compilation date 8 October 2026. Read 16 November 2026. legislation.gov.au
- "Cyber risk, Insurance Council of Australia", The insurance industry's own peak body, and the source for the statement that writing a cyber policy can require insurers to access a client's data and IT processes and potentially test their cyber defences, and that the ever-changing nature of cyber risk means coverage cannot be predicted on prior historical claims experience. Read 16 November 2026. insurancecouncil.com.au
Related Reading
- Your Board Asked About the Cyber Insurance. Here Is How to Give Them a Real Answer.
- What Underwriters Check Before They Cover an Australian Business
- How Does an Insurer Check What You Told Them?
- What Actually Happens in a Cyber Strength Audit
- The 6 Step Cyber Strength System
About the author: Kane Nawrocki is the founder and CEO of inSUPPORT. He has spent more than 25 years in IT and built inSUPPORT to give Australian businesses managed IT, security and compliance as one model, with the remediation an audit finds included in the support fee rather than billed back as a surprise project.
Content reviewed by Probably Genius for accuracy and relevance.
inSUPPORT provides managed IT and cyber security services. It is not an insurer, insurance broker or underwriter and does not hold an Australian Financial Services Licence. Where cyber insurance forms part of a plan, it is arranged through licensed insurance partners and underwritten by the insurer. Cover is subject to the insurer's assessment, the policy terms and the Product Disclosure Statement and Target Market Determination. This article is general information about IT and security practice, not financial product advice, and it does not take account of your objectives, financial situation or needs.
CLICK HERE


