Cyber security audit: what it covers and how it works

Cyber security audit explained: what it checks, how it differs from assessments and pen tests, when NIS2, DORA and GDPR expect one, and how it runs.

9 min read
Cyber security audit: what it covers and how it works

Key takeaways

  • A cyber security audit is a systematic, independent and documented check of how security is actually managed, measured against criteria agreed in advance: a standard, a law, a contract or an internal policy.
  • It differs from a security assessment, which rates posture without a formal reference, and from scans and penetration tests, which produce technical evidence but do not check processes, roles and records.
  • NIS2 Article 21(2)(f) requires policies and procedures to assess the effectiveness of risk-management measures; Articles 32 and 33 let authorities subject essential entities to regular and targeted audits, and important entities to targeted audits.
  • DORA Article 6(6) places the ICT risk management framework under regular internal audit by independent auditors with ICT risk expertise, and Article 6(7) requires a formal follow-up of findings.
  • An audit runs in sequence: objective and criteria, scope, planning, evidence gathering, evaluation, report and follow-up with corrective actions.
  • A useful report ties every finding to a requirement and a piece of evidence, rates it by severity and ends with a remediation plan with owners and dates.

A cyber security audit is a systematic, independent and documented check of how your organisation actually manages security, measured against criteria agreed before it starts: a standard such as ISO/IEC 27001, a law such as NIS2 or DORA, a customer contract or your own policies. It relies on evidence rather than statements, and ends with graded findings and a remediation plan.

The term is used for very different services, from an automated scan to a certification body's visit. This guide sets out what an audit is in the strict sense, how it differs from an assessment or a technical test, what it covers, when EU rules expect one, how it runs, and what to check before you commission one.

What an audit means in the strict sense

The reference definition comes from ISO 19011:2018, the guidance standard for auditing management systems (clause 3.1). In summary, an audit is a systematic, independent and documented process for obtaining objective evidence and evaluating it objectively, to determine the extent to which the audit criteria are fulfilled.

Three words separate an audit from an ordinary review:

  • criteria: you are checked against something written and agreed, not against the auditor's opinion;
  • evidence: records, configurations, minutes and logs, in other words facts a third party can check again;
  • independence: the person checking is not the person running the activity being checked.

Audit, assessment and testing: the difference that matters

In sales language, audit, assessment and review are often swapped. They answer different questions, and it helps to know which one you are buying.

An assessment measures, an audit judges conformity

A security assessment rates your security posture and shows where you are exposed, even without a formal reference. An audit checks conformity against stated criteria and produces findings that someone has to close. An assessment can tell you that access management is weak; an audit tells you which requirement is not met, on what evidence, and who owns the fix.

Technical tests produce evidence, they do not replace the audit

Vulnerability scanning lists known weaknesses across your estate (the tooling is compared in vulnerability management tools), while a penetration test exploits them in depth to show their impact (what a good engagement includes is in penetration testing services). Inside an audit, these tests are a source of evidence: they show whether a technical control works. They do not show whether the processes, roles and records that the criteria ask for exist.

A gap analysis comes first

A gap analysis measures how far you are from a standard before you commit to a programme, and it has no formal standing. An audit assumes the system is already running and checks that it works.

What a cyber security audit covers

The scope depends on the criteria, but the same areas come up almost every time. A useful reference is the minimum list of measures in Article 21(2) of the NIS2 Directive (Directive (EU) 2022/2555), which covers most of what an audit looks at even outside NIS2:

  • risk analysis and information system security policies;
  • incident handling;
  • business continuity, such as backup management, disaster recovery and crisis management;
  • supply chain security, including the relationship with direct suppliers and service providers;
  • security in acquisition, development and maintenance of systems, including vulnerability handling and disclosure;
  • assessing the effectiveness of the measures, which is point (f) and where the audit itself sits;
  • basic cyber hygiene and training;
  • cryptography and, where appropriate, encryption;
  • human resources security, access control and asset management;
  • multi-factor or continuous authentication, secured communications and secured emergency communication systems, where appropriate.

For each area the auditor looks for the same three things: that the rule exists, that it is applied, and that you can prove it. An approved access policy is a document; the last user access review, with its date and outcome, is evidence.

Types of cyber security audit

By who carries it out

ISO 19011 distinguishes internal audits, also called first-party audits, run by the organisation itself or on its behalf, from external audits. External audits are either second-party, run by someone with an interest in the organisation such as a customer, or third-party, run by independent bodies such as certification bodies, or by authorities.

For a mid-market company the difference is practical: you plan the internal audit, a customer's audit arrives with its own questionnaire and timetable, and a third-party audit decides a certificate or a supervisory outcome. If certification is the goal, how to pick the body is covered in ISO 27001 certification bodies.

By criteria

  • ISO/IEC 27001: the internal audit is a requirement of the standard (clause 9.2), and the certification audit is carried out by an accredited body.
  • NIS2: the risk-management measures of Article 21, as transposed into the national law of each Member State.
  • DORA, for financial entities: the ICT risk management framework.
  • GDPR: the technical and organisational security measures of Article 32.
  • Contractual requirements from a customer or a group, and internal policies.

By breadth

An audit can cover the whole management system or be targeted: one process, one site, one supplier, or one area such as identity or cloud. A targeted audit after an incident answers a precise question: which control failed, and why.

When EU rules expect an audit

There is no single obligation. Several texts ask you to verify that security measures work, each in its own words.

NIS2

  • Among the minimum measures for essential and important entities, Article 21(2)(f) lists "policies and procedures to assess the effectiveness of cybersecurity risk-management measures". Article 21(4) adds that an entity which finds it does not comply takes the necessary corrective measures without undue delay.
  • For essential entities, Article 32(2)(b) requires Member States to give competent authorities the power to subject essential entities to "regular and targeted security audits carried out by an independent body or a competent authority", plus ad hoc audits, including after a significant incident or an infringement (point (c)).
  • For important entities, Article 33(2)(b) provides for targeted security audits only, and supervision is ex post: it starts when the authority has evidence, indication or information of possible non-compliance (Article 33(1)).
  • In both cases, targeted audits are based on risk assessments, their results are made available to the authority, and the cost of a targeted audit carried out by an independent body falls on the audited entity, except in duly substantiated cases where the authority decides otherwise. Authorities can also ask for the results of security audits carried out by a qualified auditor, with the underlying evidence.

For digital providers such as cloud, managed service and managed security service providers, Commission Implementing Regulation (EU) 2024/2690 goes further. Its Annex asks for compliance monitoring at planned intervals (point 2.2) and for an independent review of network and information security, carried out by people with audit competence (point 2.3). When internal staff run the review, they are required to "not be in the line of authority of the personnel of the area under review" (point 2.3.2). The wider framework is in our guide to the NIS2 Directive.

DORA

For financial entities other than microenterprises, Article 6(6) of Regulation (EU) 2022/2554 places the ICT risk management framework under regular internal audit, in line with the entity's audit plan. The auditors need sufficient knowledge, skills and expertise in ICT risk and appropriate independence, and the frequency and focus of ICT audits are commensurate with the entity's ICT risk. Article 6(7) requires a formal follow-up process, with rules for the timely verification and remediation of critical ICT audit findings. Under Article 5(2)(f), the management body approves and periodically reviews the ICT internal audit plans and ICT audits.

GDPR

Among appropriate security measures, Article 32(1)(d) of Regulation (EU) 2016/679 lists "a process for regularly testing, assessing and evaluating the effectiveness of technical and organisational measures for ensuring the security of the processing".

ISO/IEC 27001 and customers

If you are certified or working towards certification, the internal audit is a requirement of the management system and the certification body's audit is a step towards the certificate. Second-party audits come on top, when a customer subject to NIS2 or DORA reviews the security of its suppliers. How those reviews work from the customer's side is in third-party risk management.

How a cyber security audit runs

  1. Objective and criteria. What you want to know, and against which reference: a clause, an article, a contract. Without written criteria it is an opinion, not an audit.
  2. Scope. Sites, processes, systems, suppliers and the period covered. What is left out is written down too.
  3. Planning. Who is interviewed, which documents are needed, which technical tests run, and who authorises them.
  4. Evidence gathering. Document review, interviews with the people who run the processes, direct observation, sampling of records and technical checks on configurations.
  5. Evaluation. Each piece of evidence is compared with the criteria: conforming, non-conforming, or an observation that flags a risk without breaching a requirement.
  6. Report and closing meeting. Findings go to the people who decide, with their severity and the supporting evidence.
  7. Follow-up. Corrective actions with owners and dates, and a later check that the actions closed the finding.

What the audit report should contain

  • objective, criteria and scope, including what was excluded;
  • the method: who was interviewed, what was sampled, which tests were run;
  • the findings, each tied to a requirement and to a specific piece of evidence;
  • a severity rating that separates what needs closing now from what can be planned;
  • recommendations and, after discussion with you, the agreed remediation plan;
  • a summary for senior management, who set the priorities and resources for the plan.

A report that lists problems without linking them to a criterion and to evidence does not survive the first question from a certification body, a customer or an authority: where does it say it should have been done this way?

Choosing who runs your audit

Before you sign, ask the provider five things:

  • Which criteria will they audit against, and will they put them in writing in the proposal?
  • What evidence will they collect beyond documents: interviews, record sampling, configuration checks, technical tests?
  • How independent are they from the work under review? A firm that built your controls is not the right one to audit them.
  • What the report looks like: ask for an anonymised sample and check that each finding cites a requirement and an evidence item.
  • What follow-up is included: a review of the remediation plan and a check that findings have been closed.

The effort, and so the cost, depends mainly on scope (sites, systems, suppliers), the number of criteria audited together and how much technical testing is included. A reliable estimate follows a scoping conversation.

Common mistakes

  • Buying a scan and calling it an audit. It is useful technical evidence, but it covers only part of the scope.
  • Undeclared criteria. If nobody wrote down what is being checked, the result cannot be compared with the next audit.
  • Documents instead of evidence. An approved procedure does not show that it is followed.
  • No follow-up. Findings without an owner and a date come back unchanged next time, and the next auditor sees it at once.

In short

A cyber security audit is worth as much as the criteria it is run against, the independence of whoever runs it and the evidence it collects. Scans, penetration tests and gap analyses are useful parts of the same work, but they answer other questions. NIS2, DORA, GDPR and ISO/IEC 27001 ask, in different words, for the same thing: show that the measures you declared work, and fix the ones that do not.

If you are preparing for an audit or want an independent view of where you stand, see how we work with mid-market organisations in cybersecurity services.

Frequently asked questions

It is a systematic, independent and documented check of how your organisation manages security, against criteria set before it starts, such as ISO/IEC 27001, NIS2, DORA or a customer contract. The auditor reviews documents, interviews staff, samples records and checks configurations, then delivers findings graded by severity and a remediation plan.
A security assessment rates your security posture and shows where you are exposed, even without a formal reference. An audit checks conformity against stated criteria, using evidence a third party can verify, and produces findings that someone has to close by an agreed date.
No. A penetration test, like a vulnerability scan, is a technical test: inside an audit it is a source of evidence on whether technical controls work. An audit also looks at processes, roles, records and suppliers, which a technical test does not cover.
Article 21(2)(f) of Directive (EU) 2022/2555 requires essential and important entities to have policies and procedures to assess the effectiveness of their cybersecurity risk-management measures. Competent authorities can also subject essential entities to regular and targeted security audits (Article 32), and important entities to targeted audits (Article 33), with the cost of an audit by an independent body normally paid by the audited entity.
It depends on the type. An internal audit is run by the organisation or on its behalf, by people independent of the activity under review. A second-party audit is run by someone with an interest in the organisation, such as a customer. A third-party audit is run by an independent body, such as a certification body, or by an authority.

Let's talk about your security.

A 30-minute call to understand where you stand today and tell you, with no commitment, where it pays to start.

Book a call