
Infinitas Security GmbH x Dr. Hörtkorn München GmbH
Cyber insurance can look deceptively simple from the outside. A company completes a questionnaire, confirms whether MFA is in place, whether backups exist and whether systems are patched, and the insurer uses those answers to assess the risk.
But behind those checkboxes, two companies can have very different risk quality. As the potential impact and complexity of the risk increase, the questionnaire becomes less of a final answer and more of the beginning of an underwriting conversation.
To understand how that works in practice, we spoke with Artur Bachmann of Dr. Hörtkorn München GmbH about how cyber insurers assess companies, which controls can determine insurability and how underwriting can create value beyond the policy itself.
Artur Bachmann is a cyber insurance specialist and corporate advisor at Dr. Hörtkorn München GmbH. With around 20 years of experience in insurance and more than a decade focused on cyber risk, he advises companies on understanding their exposure and finding appropriate cyber insurance solutions.
In conversation with Nicolas Inzelmann, CEO and Founder of Infinitas Security, Artur discussed why an honest "No" in an insurance questionnaire can sometimes be better than an uncertain "Yes", what companies need to maintain after placement, and where stronger cybersecurity can create room for more competitive insurance terms.
The result is a more nuanced picture than either extreme. Cyber insurance is not always a deep security audit, but it is also not simply about collecting as many positive answers as possible. Depending on the size and exposure of the company, a short checklist may be enough to obtain coverage. As complexity rises, however, insurers increasingly need to understand what sits behind the answers, and whether the company can keep that reality aligned with what it has declared over time.
For smaller companies, a short questionnaire can still be enough to obtain strong cyber coverage. Artur explains that, in some small company placements handled by Dr. Hörtkorn, short forms with only a handful of questions can be sufficient if the required criteria are met.
The reason is not that smaller companies are immune to cyber incidents. In Artur's experience, smaller businesses experience incidents more frequently than others, including business email compromise and ransomware. But the potential loss can look very different when the environment consists of only a small number of servers and clients and can sometimes be rebuilt relatively quickly. That changes the economics of the risk for the insurer and can justify a much leaner underwriting process.
The picture changes as companies and their potential losses become more complex. Especially for larger organizations, the questionnaire is much more likely to be the starting point. If an important box is not checked, the real underwriting conversation begins: Why is the measure not implemented? What is the alternative? What is the actual exposure behind it?
Revenue can help determine how deeply risks must be assessed, but it is not a complete risk model. Based on the risks he works with, Artur sees the range between roughly EUR 20 million and EUR 50 million in revenue as particularly dependent on the individual exposure, rather than as a fixed threshold across the insurance market. A hospital, for example, can represent a very different risk from another company with similar revenue. At higher revenue levels, the security controls discussed later in the interview become much more consistently expected for the SME risks he works with.
The strictness of underwriting also moves with the insurance market. Artur recalls significant cyber losses for insurers around 2022, followed by a much harder market in 2023, when premiums increased and obtaining coverage became considerably more difficult. In his current market view, premiums have fallen again, and stronger competition is now prompting some insurers to lower underwriting hurdles to win business.
What remains consistent is the focus on real loss experience. Insurers pay particular attention to common entry points for attackers and to controls that reduce the likelihood or impact of a claim. In the major claims Artur has seen in recent years, external access without MFA has repeatedly been an important entry route.
The checklist has not disappeared. What changes is how much insurers need to understand behind the answers.
Two companies may both state that they use MFA, maintain backups and operate patch management, yet the underlying risk quality can still be very different. The decisive question is often what the company actually means when it answers "Yes".
Artur uses external access as a simple example. If an insurer asks whether all external access is protected by MFA, the scope is not limited to internal employees. Supplier or service provider access can matter as well. A company may have MFA for every employee while leaving one external connection outside that protection. On paper, it may still be tempting to answer "Yes". In a claim that entered through exactly that connection, however, the scope of the original answer becomes highly relevant.
“I actually like questionnaires with more ‘No’ answers than ‘Yes’ answers. If someone isn’t sure, that ‘No’ gives us a place to start the conversation.”
Shared accounts illustrate the same point. A shared account on an isolated production machine can represent a very different exposure from a shared account with email, internet connectivity or broader system access. The checkbox alone does not describe that difference. The context does.
That is why Artur recommends accuracy over optimism. If a company is not genuinely sure that the answer is "Yes", he would rather see a "No" followed by a discussion of the context. A negative answer gives the broker and insurer something concrete to clarify: What exactly is missing? How material is the exception? Is there a compensating control? Is remediation planned?
This also puts responsibility on the insured company to understand what it is declaring. A questionnaire should not be delegated as a purely administrative exercise where one person assumes that a control "basically exists". The people answering it need to understand the scope of the question and the reality of the environment behind it.
Some weaknesses create a conversation. Others can become genuine blockers. For the SME risks discussed in the interview, Artur identifies a small group of controls where insurers can have very limited room for negotiation: MFA, immutable backups and the absence of unsupported end-of-life systems. Network segmentation between office IT and operational technology is also commonly expected where the company has a relevant production or OT exposure.
Patch management is important as well, but it can allow more discussion. A company may have a valid operational reason for not deploying a critical patch immediately, for example because it first needs to test whether the update could disrupt production. In that case, the underwriting conversation can focus on the process, the timing and the way the risk is managed rather than on a binary checkbox.
The same nuance applies to network segmentation. Its relevance depends on the company's architecture and the specific risk scenario. Where production or OT systems are part of the environment, separating them from office IT can be a critical control; where no meaningful OT environment exists, that requirement may not apply in the same way.
An unresolved weakness also does not necessarily make the entire company uninsurable. Artur gives the example of a client with an end-of-life system that remained connected because an update repeatedly caused another operational system to fail. The risk was reduced through technical constraints, but it was not eliminated. After an extended transition period, the insurer did not cancel the entire policy; it excluded the scenario associated with the unresolved end-of-life exposure.
That example also shows why remediation cannot stop at the application. Artur says he has seen requirements remain open for two or three years despite having been agreed with the insurer. If a company repeatedly fails to implement an agreed measure, the insurer can eventually stop carrying that specific exposure. In other cases, insurers may require a broader governance improvement. For example, insurers may require that an ISMS is implemented or at least that implementation is started before the next renewal.
The practical point is not that every gap must be fixed before a company can obtain cyber insurance. It is that the company needs to know which gaps are critical to coverage, which can be explained or remediated over time, and who is responsible for closing them.
It is easy to assume that stronger underwriting automatically means large evidence packages and deep technical verification. Artur describes current practice as more pragmatic.
An established ISMS is a positive signal. So is a clearly designated information security role, particularly a CISO, because it indicates that security is positioned visibly within the organization. But Artur says insurers still request relatively little formal evidence in many underwriting processes and rely heavily on the questionnaire and the risk information provided during subsequent conversations.
“Insurers actually ask for relatively little evidence. They rely heavily on the questionnaire, which makes it even more important that the information provided is correct.”
Information provided during underwriting forms part of the basis on which the insurance relationship is established. Declared controls therefore need to reflect reality and, where they are relevant to ongoing obligations, need to be maintained.
This is where documentation becomes useful even when the insurer did not request a large evidence package at placement. A temporary technical failure is different from a control that was never really implemented or has quietly disappeared. If a company can show that a measure was normally maintained, that an exception was temporary and that the issue was identified and managed, the situation can be assessed very differently.
Artur also emphasizes causality in the context of a claim. A missing or failed control is not relevant in the abstract; what matters is whether it is connected to the loss scenario under discussion. That is another reason why accurate scoping and documentation matter more than creating the appearance of perfect security.
A roadmap can add useful context even when it is not a formal coverage requirement.
“Underwriters are still people, not machines. If they see that a company knows its weaknesses and is addressing them step by step, that creates a better feeling about the risk.”
Companies react very differently to cyber insurance questionnaires. Some treat them as a useful exercise, while others mainly experience them as administrative work. The difference often lies in what happens with the findings.
For Artur, insurer questions are rarely arbitrary. They tend to focus on scenarios that insurers have actually observed and where claims experience shows a meaningful loss potential. That means the questionnaire can provide more than insurance information: it can highlight weaknesses that have repeatedly mattered in real incidents.
Backup recovery is a good example. Having a backup is only useful if the company can restore it when an incident occurs. Artur explains that insurers increasingly focus on recovery testing because failed restores have extended business interruption and increased claim costs in real incidents. A question about restore testing is therefore not simply another compliance request. It points to a very practical failure mode.
He has also seen the questionnaire become a broader improvement discussion. In one client case involving a company with more than EUR 100 million in annual revenue, Artur went through the responses and highlighted areas where comparable companies were better prepared. Not every point needed to be fixed to obtain coverage. But the process still exposed practical weaknesses and created a conversation about where the company could improve its security.
This is where the insurance assessment can become more than a gate to coverage. If a gap is identified, the company can decide whether it is critical to coverage, whether it matters to its own risk profile and how it should be addressed. The findings can then be translated into priorities, owners and a roadmap instead of disappearing once the policy has been placed.
“The cybersecurity industry has massively benefited from insurers becoming stricter, because insurers regularly require measures that matter most and companies then have to implement.”
Artur sees a wider market effect as well. Insurers learn from claims, translate that experience into underwriting requirements and, by doing so, push companies toward controls that have proven relevant in practice. His view is that the cybersecurity market has benefited from that pressure: requirements around topics such as backup recovery have made companies invest in controls that can reduce the impact of incidents.
A useful underwriting conversation therefore does not need to become a full technical audit. What helps is structured information that tells a coherent story: the important technical facts, the organizational setup, the known weaknesses and what the company is doing about them.
The discussion also touched on the value of translating technical posture into business impact. Understanding what an IT outage or other material cyber scenario could mean financially helps connect security controls to the loss exposure that the insurance is ultimately intended to absorb. For companies, that can make both security prioritization and the insurance conversation more meaningful.
And once this information exists, it should ideally be reusable. A concise summary for insurers of the company's relevant controls, organizational maturity and roadmap can reduce the need to rebuild the same risk story from scratch at every renewal. The goal is not to give the insurer a permanent live feed of the security program. It is to keep the relevant information understandable, current and ready when it is needed.
What a company declares during underwriting cannot simply be treated as a permanent snapshot if the underlying environment changes. Systems are replaced, suppliers are added, responsibilities move, exceptions appear and controls can drift. Getting coverage at one point in time therefore does not remove the need to keep the declared security posture aligned with reality.
Artur stresses that agreed measures need to be maintained. He also sees value in someone deliberately revisiting the risk on a regular basis. At Dr. Hörtkorn, the usual approach is active rather than passive: many placements are based on annual contracts that renew automatically if they are not cancelled, while the risk is nevertheless reassessed each year through re-underwriting and an updated questionnaire. He also notes that not every insurer follows the same process.
“Someone who reassesses the risk regularly and restarts the process helps make sure the company remains insurable in the future.”
The annual review serves more than one purpose. It gives the company a chance to catch the gap between what was originally declared and what now exists. It allows the broker and insurer to reflect new loss experience and changing underwriting priorities. And it keeps the company ready for the wider insurance market.
That last point is easy to underestimate. An insurer can change its appetite, leave a market or simply decide that it no longer wants a particular risk. If the company has not revisited its security position for several years, it may suddenly need replacement coverage while also discovering that it is no longer ready for the insurance market. At that point, remediation becomes urgent and negotiating leverage declines.
Regular reassessment can work in the opposite direction as well. If a company has materially improved its risk, it can bring that progress back to the broker proactively rather than waiting for the next problem. Artur gives examples such as a properly implemented and tested business continuity setup or a substantially improved reporting and management process. Meaningful improvements can create a reason to revisit the market or ask whether better conditions are available.
The point is not continuous monitoring for its own sake. It is maintaining a sufficiently current picture of risks, controls, responsibilities and improvement activity so that the company remains both secure and insurable as its environment changes.
The value of continuous cyber management becomes particularly visible once an incident occurs. Insurance can provide access to specialists and financial protection, but the speed and quality of the response also depend on what the company has prepared internally.
“Cyber claims are handled much better, and much more cheaply, when a company has proper business continuity management in place.”
A BCM setup that is properly maintained can reduce the time between impact and the first effective response measures because responsibilities, escalation paths, contact details and supporting service providers are already known. That can help reduce downtime and make the response more controlled.
The word "maintained" matters. A business continuity plan that still contains old phone numbers, outdated responsibilities or former service providers is much less useful in a crisis. BCM therefore illustrates the wider difference between having documentation and operating a living security process.
Artur does not currently describe BCM as a universal underwriting deal breaker in the same category as MFA or immutable backups. But he considers it highly regarded by insurers because it can materially improve the way a claim is handled and reduce the operational impact of an incident.
This is also where insurance and continuous security management complement each other rather than compete. The policy provides an external response ecosystem and financial protection. Internal preparedness helps the company use that support faster and with better information.
“Traditionally, we didn’t talk about a measure significantly reducing your premium. We talked about that measure making you insurable. That is beginning to change.”
Not in a simple direct relationship. For a long time, Artur says, the primary insurance benefit of implementing required security measures was simply becoming insurable: the measure is in place and coverage is possible, or the measure is missing and coverage may not be available.
That is beginning to change as insurers compete more strongly for attractive risks. In competitive situations, a very good security setup can create additional room on pricing. Artur describes this as an opportunity to gain the last few percentage points in the premium rather than a fixed discount for every additional control.
The distinction matters. Installing one more security tool does not automatically produce a measurable premium reduction. But a materially better risk profile, especially one that is clearly documented and can be explained to the market, can strengthen the company's position when insurers compete for the business.
That creates a useful hierarchy for security investment. First, the company needs to meet the requirements that make the risk insurable. Second, it needs to keep those controls effective so that its declarations remain credible. Third, broader improvements in resilience and management can reduce the actual loss exposure and, in the right market situation, may also support better insurance conditions.
In other words, stronger security should not be justified only by a possible premium discount. The more important value is lower operational risk and continued access to coverage. Better pricing can be an additional benefit.
Artur closes with three recommendations that are deliberately practical.
Cyber insurance is a complex product and differences in conditions can go into considerable detail. A specialist should understand both the insurance wording and enough of the client's IT and risk to explain what the insurer is actually asking and why.
That translation role matters because some questions can appear unrelated to cybersecurity at first glance. Artur gives the example of questions about copyright checks before publishing media content: they make sense once the company understands that media liability can form part of the cyber coverage. A specialist broker should be able to explain the risk behind questions like these instead of acting as a messenger who simply forwards a form.
“If you choose a broker who does not specialize in cyber, you risk having little more than a messenger between the insurer and the company, someone who doesn’t understand the content.”
If the organization is not genuinely sure that an answer is "Yes", Artur recommends choosing "No" and discussing the detail. The objective is not to maximize positive answers. It is to make sure the information accurately reflects the real environment.
A "No" does not automatically mean that the risk is uninsurable. It can open the conversation about scope, compensating controls and remediation. An uncertain "Yes" may feel easier during the application, but it creates a much more difficult situation if the declared control later becomes relevant to a claim.
IT and information security leads, together with the relevant operational teams, need to understand the technical obligations arising from the policy and the security posture declared in the questionnaire. If the process is handled primarily on the commercial side, those requirements still need to reach the people who operate the controls every day.
That includes what has been promised, which controls need to remain in place and what needs to be remediated. A policy cannot remain aligned with the technical reality if the people managing that reality do not know what the company declared.
There is also a useful fourth principle underneath all three recommendations: understand why the insurer is asking the question. The questionnaire is much more valuable when the company sees the loss scenario behind the checkbox rather than treating the form as a test to be passed.
Taken together, the interview points in one direction: do not treat the cyber insurance questionnaire as an administrative exercise that happens only once. Treat it as a conversation about risk, make sure the answers reflect reality and use the findings to continuously improve the company’s security maturity rather than simply to obtain the policy.
Ultimately, cyber insurability is not about presenting a perfect security picture. It is about understanding what insurers are asking, being accurate about the controls that are actually in place and showing that known gaps are being managed. A continuously maintained security posture can improve resilience, simplify future underwriting and keep the company ready if the insurance market or its own risk profile changes.
Becoming insurable is one thing. Remaining insurable is the ongoing challenge.
Read more about what changes when a cyber incident actually happens and explore how insurance, internal readiness and continuous cyber management can shape response, recovery and what happens afterwards. Read the full article here.
What determines whether a company is eligible for cyber insurance?
What cybersecurity controls are most important for cyber insurance?
Can stronger cybersecurity reduce cyber insurance premiums?

Compare three security approaches before, during and after a cyber incident, and see how insurance readiness shapes response, recovery and improvement.
Mehr lesenCyber insurance can look deceptively simple from the outside. A company completes a questionnaire, and the insurer uses those answers to assess the risk. But behind those checkboxes, two companies can have very different risk quality. To understand how that works in practice, we spoke with Artur Bachmann of Dr. Hörtkorn München GmbH about how cyber insurers assess companies, which controls can determine insurability and how underwriting can create value beyond the policy itself.
Mehr lesen