Security
How Often Should You Conduct Penetration Testing?
8 Sept 2026 · Updated 8 Sept 2026
Annual penetration testing is the minimum for most organisations, but higher-risk environments and those with active change programmes need more frequent assessments. PCI DSS mandates at least annual testing plus a test after significant changes; the ASD Information Security Manual recommends annual testing for systems handling sensitive information. A fixed annual cadence combined with trigger-based testing for major changes is the practical baseline.
Penetration testing is a point-in-time assessment. Its findings reflect an organisation’s security posture at that moment, and the moment passes quickly as systems change, new vulnerabilities emerge and attackers refine their techniques. Getting the frequency right matters as much as the test itself.
The baseline: annual testing
For most organisations, annual penetration testing is the practical and regulatory minimum. This cadence is well-supported by the major Australian and international frameworks. The ASD Information Security Manual recommends annual penetration testing for systems handling sensitive information, and more frequent testing for higher-risk environments. PCI DSS v4.0.1 requires testing at least annually under requirements 11.4.2 for internal testing and 11.4.3 for external testing. ISO 27001 expects periodic technical vulnerability testing as part of a risk-based control programme, with frequency determined by the risk assessment. SOC 2 has no explicit mandate but auditors routinely expect a penetration test to fall within the 12-month audit period.
Annual testing works well when a system is relatively stable, changes are managed through a formal process, and the risk profile is moderate. It gives a verifiable baseline, satisfies most regulatory requirements, and provides a meaningful cycle for remediation and re-test.
When to test more frequently
Annual testing is not sufficient for all environments. Several factors push the cadence to every six months or to continuous trigger-based testing.
High-risk environments. Systems that process payment card data, health information, financial transactions or government classified information face greater threat attention and stricter regulatory expectations. Organisations subject to the Notifiable Data Breaches scheme under the Privacy Act have a direct incentive to discover and remediate vulnerabilities before an attacker does, not after.
Rapid development and change. Cloud-native applications and teams using agile or continuous delivery may release code weekly or daily. A single annual test cannot keep pace with that rate of change. Many organisations embed lightweight security testing into the development pipeline alongside an annual or biannual full-scope test.
Post-incident or threat intelligence signals. A security incident, a near-miss, or intelligence that the organisation’s industry is actively targeted are all triggers for an out-of-schedule test. Waiting 11 months after a breach to test again is not defensible.
Expanding attack surface. Adding a new internet-facing service, deploying a new API, connecting a third-party integration, or migrating workloads to cloud all introduce new attack surface that was not present in the last test. Each is a trigger.
Mergers and acquisitions. Acquiring a company means inheriting its security posture. A penetration test of the acquired environment before integration is standard practice for large Australian organisations.
Framework-specific requirements
| Framework | Requirement |
|---|---|
| PCI DSS v4.0.1 | Annual testing; additional test after significant change (req. 11.4.2 and 11.4.3) |
| ASD ISM | Annual penetration testing recommended for systems handling sensitive information |
| ISO 27001 | Frequency set by risk assessment; periodic technical vulnerability testing |
| SOC 2 | No explicit mandate; auditors typically expect a test within the audit period |
| Essential Eight | No frequency prescribed; controls should be regularly verified as effective |
Where multiple frameworks apply, the most frequent requirement governs. A PCI DSS-scoped organisation that also seeks ISO 27001 certification runs at least annual testing to satisfy both.
Trigger-based testing: what should prompt an out-of-cycle test?
Beyond the regular cadence, specific events should trigger an additional test regardless of when the last one occurred.
A major infrastructure change, new internet-facing systems, a significant update to authentication architecture, moving from on-premises to cloud, a network segmentation change, and adding or removing a third-party integration all create new attack paths that the previous test did not examine. A merger or acquisition where the acquired IT environment is being connected to the existing one is high on the list. And an actual security incident, or credible threat intelligence about the sector, warrants a fast-follow test once the immediate response is complete.
The practical approach is to document these triggers formally, so the decision to test is not discretionary when one occurs.
Re-testing after remediation
A penetration test is not complete when the report is delivered. Findings need to be remediated, and remediation needs to be verified. Most organisations target remediation of critical and high findings within 30 days and medium findings within 90 days, with a re-test of remediated items to confirm the fix held and did not introduce a new weakness. A re-test is usually a scoped, targeted engagement focused on the original findings rather than a full repeat of the original test.
Skipping the re-test is a common mistake. A patch applied in a hurry may fix one issue but leave a related vulnerability in place, and the only way to know is to test again.
What affects cost and scope?
Penetration test frequency decisions are partly budget decisions, so it helps to understand what drives cost. Scope is the primary variable: the more systems, applications and network segments in scope, the longer and more expensive the test. An external perimeter test of an organisation’s internet-facing systems is narrower and cheaper than a full-scope test covering external, internal network, web applications and wireless. Scoping tests to the highest-risk areas, particularly before the annual full-scope test, is a reasonable cost management approach.
Vulnerability scans are much cheaper than penetration tests and can run more frequently as a between-cycle check. They are not a substitute for a test, but they can catch newly disclosed vulnerabilities before the next formal engagement.
Getting started
For an organisation not currently on a testing cadence, the place to start is its risk profile and regulatory obligations. PCI DSS, the ASD ISM or an industry regulator may set a minimum. If none of those apply, annual full-scope testing with a trigger-based policy for major changes is a sound baseline for most Australian organisations. Cybernion offers penetration testing services scoped to the environment and risk profile.
Frequently asked questions
How often should we do penetration testing?
For most organisations, annual penetration testing is the minimum. Higher-risk environments, those that handle sensitive data, face regulatory scrutiny or change rapidly, typically need testing every six months or after every significant change. PCI DSS mandates at least annual testing plus a test after significant changes. The ASD Information Security Manual recommends annual penetration testing of systems handling sensitive information. A fixed annual test plus a trigger-based test for major changes is the practical approach for most Australian businesses.
What triggers a penetration test outside the regular cycle?
Major infrastructure or application changes, a security incident or near-miss, a merger or acquisition, adding a new internet-facing service, a significant change to the authentication or network architecture, and moving workloads to cloud all warrant an out-of-cycle test. Regulatory requirements may also require testing after significant change. Waiting for the annual test after a major change leaves an unverified gap in the organisation’s security posture.
Does penetration testing frequency differ by framework?
Yes. PCI DSS v4.0.1 requires testing at least annually and after significant changes (requirements 11.4.2 for internal testing and 11.4.3 for external testing). The ASD Information Security Manual recommends annual testing for systems handling sensitive or classified information. ISO 27001 leaves frequency to the risk assessment but expects periodic testing. SOC 2 has no explicit mandate but auditors typically expect a penetration test within the audit period. Essential Eight does not prescribe testing frequency but expects controls to be verified as effective.
How long is a penetration test valid?
A penetration test is a point-in-time assessment. Its findings are valid until a significant change occurs, a new vulnerability is discovered in the tested systems, or the agreed assessment period lapses, usually 12 months for most frameworks. Using a test result older than 12 months to assert security assurance is generally not accepted by auditors or regulators. Re-test remediated findings within 30 to 90 days to confirm fixes held.
Should small businesses do penetration testing?
Small businesses with limited budgets often prioritise a vulnerability scan first, but any organisation that handles financial data, sensitive personal information, or has internet-facing systems benefits from at least an annual penetration test. The cost of a test is typically far less than the cost of a breach. Scoped, targeted tests for a specific application or external perimeter are cheaper than a full-scope engagement and appropriate for smaller organisations.
Sources:
- PCI DSS v4.0.1, PCI Security Standards Council, 2024
- Information Security Manual, Australian Signals Directorate, 2026
- NIST SP 800-115, Technical Guide to Information Security Testing, NIST Computer Security Resource Center, 2008
Last updated: 30 August, 2026
