
DORA Audit Fire Drill: The Dress Rehearsal Before the Audit

Policies updated, responsibilities assigned, controls established: In many financial institutions, implementation of the Digital Operational Resilience Act (DORA) is thus considered to be well underway and, in many cases, formally implemented. But the real test begins only when an auditor asks not about the concept, but about how the controls are actually implemented. Were they carried out at the scheduled intervals? Do they truly cover all relevant systems, applications, and users? And can their effectiveness be reliably demonstrated over the entire audit period?
It is precisely at these junctures that unpleasant surprises arise in practice. A process is documented but does not cover the entire system landscape. A control is performed, but not at the required frequency. An exception was approved without its compensating measure being subject to mandatory monitoring. Or the technical implementation works, but cannot be verifiably demonstrated due to a lack of evidence.
A DORA audit fire drill deliberately brings this test to the forefront. As a realistic audit simulation, it tracks selected controls from the regulatory requirement through the process and technical implementation all the way to the supporting evidence. This ensures that critical weaknesses do not first become apparent during the actual audit, but rather at a time when they can still be prioritized, remedied, and retested.
DORA Readiness: Prevention Creates Room for Maneuver
For many financial institutions, DORA compliance has now been formally completed — but a closer look reveals weaknesses in the control and evidence chain that were not identified internally. Therefore, a structured assessment of a company’s own compliance status provides early transparency regarding existing gaps and helps address key issues in a timely manner. To this end, Horn & Company has developed a DORA maturity model that systematically maps more than 50 DORA-relevant capabilities and rates them across five maturity levels, ranging from “rudimentary” to “optimized.”
The model highlights the areas where the current maturity level still deviates significantly from the target state. It does not merely examine individual controls but evaluates the capabilities relevant to DORA in terms of their organizational and technological aspects. This allows areas for action to be prioritized early on and measures to be initiated before existing gaps could lead to potentially serious findings during an audit. A high level of readiness thus creates room for maneuver early on and reduces the need for last-minute, extensive audit preparations.
From a Readiness Check to a Fire Drill—What's the Difference?
The term “DORA Audit Fire Drill” does not refer to a regulatory-defined type of test. We define it as a voluntary, time-limited, and risk-based audit simulation conducted prior to an internal, external, or regulatory audit. This simulation thus complements the DORA Readiness Check but does not replace either the digital operational resilience tests required under DORA or a formal independent audit.
While the Readiness Check provides transparency regarding the implementation status and maturity level of the relevant capabilities, the Fire Drill goes a step further: It deliberately places selected controls in a realistic audit scenario. The key question then is: Can the organization demonstrate, based on concrete cases and robust evidence, that a control actually works?
What Makes a Fire Drill Effective
For the fire drill to go beyond a preparatory series of interviews, three closely interlinked components must work together:
- Audit Logic: The scope is not understood as an unweighted, comprehensive audit of all DORA requirements. The criticality of business processes and systems, previous findings, technical complexity, and the current maturity level of the respective control areas — assessed, for example, using the DORA maturity model from Horn & Company — determine where an in-depth test is warranted. In particular, areas with low control maturity and an increased potential for significant or serious findings should be prioritized early on.
- Technical validation: Control descriptions are traced through to the actual system environment. Depending on the audit area, this may include, for example, data from Identity and Access Management (IAM) and Privileged Access Management (PAM), system and user populations, database or operating system parameters, firewall rules, logs, tickets, exception approvals, and historical control results.
- Remediation and Retest: Findings are prioritized based on severity and the need for implementation; short-term risks are mitigated immediately, and structural measures are managed through the process of re-executing controls and retesting.
It is only the interplay of these three components that transforms audit preparation into true audit readiness.
What's Put to the Test in a Fire Drill
Which areas are tested in greater depth depends on the organization’s risk profile and its system landscape. The scope ranges from governance and second line of defense topics — such as DOR strategy, policies, and control frameworks — to technical audit areas such as logical access protection, privileged access, recertifications, and network and resilience issues. Self-managed applications, data retention, or the geographic redundancy of critical infrastructure can also be part of a fire drill.
What matters most is not so much the individual audit area as the nature of the weakness that becomes apparent:
- Design Gap: The control exists but does not adequately address the underlying risk — for example, in the case of segregation of duties (SoD), where critical authorization conflicts can be approved without requiring an effective compensating control.
- Coverage Gap: The control functions in principle but does not cover all relevant systems, accounts, or applications — for example, when individual database platforms are still outside the scope of a PAM scan.
- Execution Gap: The control is not executed, or is not executed with the required frequency — for example, when certain firewall rules are reviewed only annually instead of also semi-annually.
- Evidence Gap: A control action may have been performed, but cannot be reliably verified due to missing tickets, timestamps, or documented decisions.
- Governance Gap: Scopes, responsibilities, or exception processes are not clearly defined.
- Structural Gap: The root cause lies deeper within the technical architecture and cannot be resolved by adjusting policies or processes.
This range of issues was also evident in an anonymized audit portfolio: The findings ranged from insufficiently managed segregation-of-duties exceptions to gaps in PAM coverage and a lack of evidence for firewall reviews, as well as ambiguities in change management and structural issues related to geo-redundancy.
Not every finding carries the same weight
That is precisely why prioritization is crucial. BaFin classifies audit findings on a scale from F1 to F4 — as the severity increases, so do the urgency and the need for action.
In a fire drill, this logic can be used to categorize potential findings based on risk. The formal classification during the actual audit is left to the respective auditor.
A well-designed fire drill, therefore, does not attempt to identify as many minor deviations as possible. Instead, it focuses in particular on control weaknesses that could lead to significant or serious findings—as well as on issues whose permanent resolution requires a long lead time.
This is because the number of findings says little about the actual need for action. A single structural problem in an authorization model, a PAM landscape, or the infrastructure can tie up more time, budget, and management attention than numerous minor documentary discrepancies.
For management, therefore, it is not only crucial how serious a weakness is, but also how long it will take to resolve it. The later a structural issue comes to light, the greater the short-term budgetary requirements, management attention, and technical implementation pressure will be.
Practical Example: When the exception process does not control the actual risk
The SoD example mentioned earlier illustrates how such a design gap can lead to a significant finding.
The identity and access management system detects a critical combination of permissions. The responsible supervisor reviews the conflict, documents a justification, and approves a temporary exception. The intended workflow is thus formally completed.
In a fire drill, however, the crucial question arises: What specific compensating control was implemented, who monitors it, and how can its effectiveness be demonstrated?
In an anonymized audit case, it was precisely this second level of control that was missing. SoD conflicts could be detected and exceptions approved without requiring additional monitoring measures or verifying that documented mitigation measures were actually being implemented. This weakness was classified as an F3 finding.
The problem, therefore, was not the absence of an IAM workflow. On the contrary: the process existed and was being used. However, its design did not ensure that the risk arising from the exception was effectively controlled. In this regard, DORA goes significantly further than previous regulatory frameworks.
A fire drill, therefore, does not merely examine individual capabilities. In particular, it also investigates whether the effectiveness of the established measures can be demonstrated based on evidence.
"Being DORA-ready doesn't mean being able to demonstrate the process. Being DORA-ready means being able to prove the effectiveness of the control."
Procedural Model: From Test Hypothesis to Retest
The process should be carried out in a focused and step-by-step manner:
1. Define the scope and audit hypotheses: The starting point consists of critical business services, relevant systems, previous findings, and known vulnerabilities. In addition, there are issues whose resolution requires a particularly long lead time. In the SoD example, the hypothesis is: The existing exception process does not reliably prevent critical authorization combinations from persisting without effective compensation.
2. Request evidence and conduct walkthroughs: A realistic request list includes policies and control descriptions as well as the complete exception population, conflict matrices, approvals, extensions, and evidence of controls. The process is then reviewed based on specific cases — not solely on the basis of its target description.
3. Conduct spot checks and technical deep dives: Before conducting spot checks, it must be assessed whether the underlying population is complete. Only then are risk-based cases selected and tracked through to their technical implementation.
4. Evaluate and prioritize findings: Severity serves as the central basis for prioritizing corrective actions. The time required for resolution, technical dependencies, existing compensating controls, and the time until the next control execution also play a significant role in determining the operational implementation sequence.
5. Resolve and retest: In the SoD example, exceptions that are no longer necessary in the short term could be closed first, and particularly critical conflicts could be reviewed. In the long term, however, clear exclusion rules, specific requirements for compensating controls, limited exception periods, and robust monitoring must be established. The retest verifies whether the new rules are effective in practice and generate reliable evidence.
Where the fire drill delivers the greatest value
Not every financial institution needs to immediately test the entire DORA framework in its full depth. Often, a focused approach is more effective.
A “fire drill” is particularly relevant when an internal, external, or regulatory audit is imminent; when new controls must be documented for a full period for the first time; or when heterogeneous system landscapes, numerous manual handoffs, and many exceptions make it difficult to provide sufficient evidence.
Previous audit findings or control areas with an increased risk of F3/F4 findings are also suitable for a targeted deep dive.
The following principle applies: Not every control area needs to be examined with the same level of depth. A risk-based scope makes it possible to cast the fire drill’s net wide enough while simultaneously delving into technical details where significant vulnerabilities or increased audit risks are expected.
The Right Time: Early Enough to Make a Difference, Late Enough to Provide Evidence
A fire drill should not be initiated only when an audit is already imminent. At that point, it is often no longer possible to resolve significant structural weaknesses in a sustainable manner — especially when technical adjustments, new control processes, or changes to the system landscape are required.
Rather, the right time is when the key controls have already been implemented and established in day-to-day operations, but there is still sufficient time to identify, remedy, and retest any weaknesses. This allows the fire drill to fulfill its actual purpose: not to catch up on DORA implementation, but to test its resilience under realistic audit conditions.
A focused fire drill can often be conducted within a few weeks. If multiple control areas, subsidiaries, or heterogeneous system landscapes are involved, the effort required increases accordingly. However, what matters most is not so much the sheer duration of the drill as the subsequent scope for action.
Insufficient evidence of controls can sometimes be improved at short notice. In contrast, missing historical evidence often cannot be generated retroactively. Structural measures — such as the full integration of privileged accounts into a PAM solution, the revision of an organization-wide authorization model, or changes to the infrastructure — require significantly more lead time anyway.
The dry run should therefore take place several months before an expected audit. The more severe a finding is expected to be and the longer the necessary remediation will take, the earlier the corresponding control area should be tested.
The internal effort involved also depends less on the number of regulatory requirements than on the complexity of the control chains under consideration. Many systems, manual handoffs, decentralized applications, and numerous exceptions increase the effort required to generate complete data sets and reliable evidence in particular.
A fire drill is typically carried out by a small, interdisciplinary core team of audit, cybersecurity, and technology specialists. On the client side, central coordination is required, as well as the targeted involvement of the respective control and system managers.
The fire drill itself typically lasts three to four weeks. The resulting remediation portfolio, however, can span several months and therefore requires early prioritization and management.
What Should Be on the Table After the Fire Drill
The result, therefore, is not just another unweighted list of deficiencies. Rather, a good fire drill provides a prioritized view of audit readiness:
Which control chains are already robust? Where is there a risk of significant or serious findings? What short-term measures can immediately reduce the risk? Which structural changes require a longer lead time? What evidence is still missing? And which controls need to be retested before the actual audit?
This transforms a multitude of individual technical findings into a manageable remediation agenda. This is precisely where Horn & Company comes in: We combine audit logic with cyber and technology expertise and support our clients from the selection of relevant control areas through technical review and prioritization to implementation. In doing so, we coordinate business, IT, and control functions and help transform identified vulnerabilities into technically and organizationally robust measures.
The formal, independent audit opinion is intentionally left to Internal Audit, the external auditor, or the regulatory authority. Our primary contribution is to identify critical weaknesses early on and work with the client to effectively address them.
Put DORA Audit Readiness Through Its Paces
Whether it’s an upcoming audit, known weaknesses, or uncertainty about the actual effectiveness of existing controls: In a no-obligation initial consultation, we’ll determine which control areas should be examined in greater depth, what scope makes sense, and how much lead time is needed for a fire drill, remediation, and retest.




