Penetration Testing Requirements: What HIPAA, ISO 27001, SOC 2 and Cyber Insurance Actually Require
Penetration testing requirements differ by framework. Only PCI DSS and the FTC Safeguards Rule name a penetration test outright, while SOC 2, HIPAA, ISO 27001 and cyber insurance get the same evidence indirectly, through auditors and underwriters who ask for a dated report. The stakes are measurable: IBM's 2025 Cost of a Data Breach Report puts the global average breach cost at $4.44 million [1]. Clone Systems' automated external penetration tests start at $1,995.
You are filling out a renewal form, and one line asks whether you conduct penetration testing. You have a clean vulnerability scan from last month and a pen test report from 14 months ago. Which answer is true? In scoping calls, we see this exact moment more than any other: the team has done real security work, but cannot say which framework is asking, how often, and what evidence it expects.
That uncertainty has a cost. In IBM's 2025 Cost of a Data Breach Report, the mean time to identify and contain a breach was 241 days [1]. And HHS has pushed the HIPAA Security Rule final rule into 2027, so covered entities stay on the current rule's risk analysis obligations in the meantime [4]. The frameworks below are the ones that drive pen tests in our scoping calls most often, and each one reads the same report differently. This post maps what each actually requires, how often, and what the evidence has to show.
Which Frameworks Actually Require a Penetration Test?
Two frameworks name a penetration test outright: PCI DSS and the FTC Safeguards Rule. The rest get the same evidence through an intermediary, and the intermediary is who you are really answering to: an auditor for SOC 2, HHS's risk analysis expectations for HIPAA, an ISMS management review for ISO 27001, and an underwriter for cyber insurance.
| Framework | Requirement anchor | What it asks for | Frequency |
|---|---|---|---|
| PCI DSS | 11.4.2 (internal), 11.4.3 (external), 11.4.4 (retest), 11.4.5 (segmentation) | Internal and external penetration tests with retest verification of fixes | At least every 12 months and after significant changes |
| FTC Safeguards Rule | 16 CFR 314.4(j) (risk assessment) | Testing of the information security program, which FTC guidance lists as including periodic penetration testing [2] | Ongoing; no fixed frequency |
| SOC 2 | TSC CC4.1, CC7.1, CC7.2 [7] | A dated external penetration test report with a Letter of Attestation, per auditor practice | At least annually, inside the Type 2 audit period |
| HIPAA | 45 CFR 164.308(a)(1)(ii)(A) (risk analysis) | No named test; a risk analysis of your systems that testing can validate | No fixed frequency; annual is the norm |
| ISO 27001 | A.8.8 (technical vulnerability management) [5] | No named test; evidence that vulnerabilities are discovered and assessed, reviewed in the ISMS | Per your ISMS policy; annual review is typical |
| Cyber insurance | Carrier underwriting [6] | A dated, independent penetration test report | Typically within the last 12 months |
Read the matrix as a chain of custody. Where a framework names a test, the requirement binds your program directly. Where it does not, the requirement binds the person who checks your evidence, and that person has their own habits. A SOC 2 auditor who has never accepted a 13-month-old report turns a "not required" line into a 12-month deadline in practice.
How Often Does Each Framework Expect a Penetration Test?
The shortest honest answer is: as often as the strictest framework you face, and never less than after significant changes. Here is how the frequencies break down.
PCI DSS: every 12 months, plus after significant changes
Requirement 11.4.2 covers internal testing and 11.4.3 covers external testing, both at least every 12 months and after significant infrastructure or application changes. 11.4.4 requires retesting to verify fixes, and 11.4.5 requires segmentation testing at least every 12 months if segmentation is your scope-reduction strategy [9]. The full requirement breakdown is in our PCI DSS requirement 11.4 explainer.
SOC 2: at least annually, inside the audit period
The Trust Services Criteria do not name a penetration test, but CC4.1, CC7.1 and CC7.2 create the expectation, and auditors have settled on a practical reading: at least one external penetration test a year, with an internal test added when infrastructure is in scope. For a Type 2 report, the test, the fixes and the retest all need to fall inside the audit period, which is why most teams run the test early in the period [7].
HIPAA: no fixed frequency, driven by the risk analysis
The Security Rule requires a risk analysis of the systems that hold ePHI, and a penetration test is one of the strongest ways to test the assumptions in that analysis [4]. Because the rule sets no cadence, covered entities in practice align the test to their annual risk analysis cycle, and the final rule HHS pushed into 2027 will not change that in the short term [4].
ISO 27001: whatever your ISMS policy says, checked at annual review
Control A.8.8 requires technical vulnerability management: discover, assess and report vulnerabilities, with a distinction between internal and external sources [5]. A penetration test is the standard external source, and your ISMS documentation decides the cadence. The annual management review is where a missing test shows up as a gap.
Cyber insurance: within the last 12 months, at renewal
Carrier forms vary, but the pattern is consistent: testing within the past 12 months, done by an outside party, with the kinds of test specified [6][8]. Renewal season is when a lapsed report stops being a paperwork problem and becomes a coverage problem.
If you face more than one of these at once, the frequency question becomes a scheduling problem. Our post on how often you should penetration test walks through building that calendar, including the retest nobody books.
Does a Vulnerability Scan Satisfy a Penetration Testing Requirement?
No, not for any line that names a test, and not for the evidence consumers that treat the two as different inputs. A vulnerability scan lists known weaknesses. A penetration test attempts to exploit them and shows how far an attacker could get, which is the proof of impact a scan cannot produce.
The distinction matters for compliance because the question is asked separately. Cyber insurance forms ask about penetration testing and vulnerability scanning as separate questions, and an answer that confuses the two becomes a misstatement once the policy is in force [8]. PCI DSS treats them as separate requirements in the first place: 11.3.2 for vulnerability scanning and 11.4 for penetration testing, which is why a clean ASV scan never counts as your 11.4 evidence. Our breakdown of penetration testing versus vulnerability scanning covers the division of labor in detail.
In our engagements, clients are often surprised to learn that the answer they gave an insurer about testing in one year was built from a scan, and that the carrier now lists them as untested at renewal. The two work well together. One is the other's answer key. But neither substitutes for the other on a form.
What Evidence Does Each Framework's Checker Actually Need to See?
The framework tells you what to do. The evidence tells the checker that you did it. Before a report goes to an auditor, an underwriter or a management review, it has to survive three checks, which we call The 3-Question Evidence Check:
- Date. Does the report's date fall inside the window the checker cares about? That is the last 12 months for PCI DSS and most carriers, the audit period for a SOC 2 Type 2, and "recent enough" for an ISO management review, which in practice means inside the review cycle.
- Scope match. Does the report name the assets, test types and access levels, so the checker can match what was tested against what the framework says should be covered? A report that says "external network" without IP ranges cannot answer a question about a new application.
- Closure. Does the report show the findings were remediated and retested? PCI DSS 11.4.4 makes this explicit, and SOC 2 auditors ask for it as a matter of course [7][9].
When we deliver a report for a SOC 2 audit, the most common follow-up from the auditor is not about the findings. It is about the retest: proof that the reported issues were fixed and verified inside the period. A report that answers the third question is the one that gets filed. If you are reading an existing report and unsure whether it would survive these three checks, our post on how to read a penetration test report gives you the reading order to use.
Test your own file right now. Pull your most recent penetration test report and run it against the three questions above. If it fails the date question, your next renewal answer is already at risk. If it fails the scope question, it may not cover the systems your framework actually asks about. If it fails the closure question, the findings it proved are still open, and the checker will eventually find out.
How Do You Keep the Answer Yes at Every Renewal?
Set the cadence to your strictest framework, not your average one. If PCI DSS is on your list, that is at least every 12 months plus a test after significant changes, which in practice means most organizations run external tests every six months to keep the 12-month guarantee honest across change events. If cyber insurance renewal is in February and your last test was in March, the answer is "yes", but only because you booked the retest.
Three moves keep the file audit-ready year round:
- Align the test calendar to the strictest deadline you face, and put the retest in the calendar with the test, not after it.
- Scope the test to the decision it supports. An insurance renewal needs a dated report on your current external surface, and a SOC 2 file needs the same report inside the audit period with an internal test when infrastructure is in scope.
- Keep the report pack with the fixes: the dated report, the Letter of Attestation, the remediation evidence and the retest result, filed where the next person can find them.
Pricing is usually where this plan stalls, so it is worth knowing the numbers before you decide. The penetration test cost guide breaks down what moves the number, and Clone Systems' automated external penetration tests start at $1,995 for one asset, with automated internal tests from $2,995 for up to 25 hosts and authenticated web application tests from $2,495.
How Clone Systems Can Help
Clone Systems runs penetration testing in three shapes: automated external and internal tests you can start online today, authenticated web application tests, and expert-led manual tests scoped to complex business logic, APIs or cloud configurations. Every test ends with a dated report pack: an executive summary, a technical report with proof of exploit, and a free retest inside your 30-day window, which is what makes the evidence survive the three questions above.
For each framework, we keep a dedicated page that maps the test to the evidence: penetration testing for SOC 2, penetration testing for HIPAA, penetration testing for ISO 27001, penetration testing for cyber insurance, and PCI DSS requirement 11.4. Comparing platforms? See how Clone Systems compares with Aikido or Rapid7, and use the penetration testing buyer guide when you are evaluating more than one vendor. For the full service, see managed penetration testing services.
Frequently Asked Questions
Is a penetration test required for SOC 2? The Trust Services Criteria do not name a penetration test, but CC4.1, CC7.1 and CC7.2 create the expectation, and auditors ask for a dated external report as evidence [7]. For a Type 2, the test, the fixes and the retest all need to fall inside the audit period.
Does HIPAA require penetration testing? HIPAA does not name a penetration test. It requires a risk analysis of your systems under 45 CFR 164.308(a)(1)(ii)(A), and a penetration test is one of the strongest ways to validate that analysis [4].
How often does the FTC Safeguards Rule require a penetration test? The Safeguards Rule requires an ongoing risk assessment of your information security program under 16 CFR 314.4(j), and FTC guidance lists periodic penetration testing as part of testing the program [2]. The rule sets no fixed frequency, so your risk assessment decides the cadence.
Does cyber insurance require a penetration test? Most carriers ask whether you conduct penetration testing, how often, and whether it is done by an outside party, and generally expect testing within the last 12 months [6][8]. A dated, independent report is what makes the "yes" on the form defensible.
What is the difference between a vulnerability scan and a penetration test? A vulnerability scan lists known weaknesses, while a penetration test attempts to exploit them to show how far an attacker could get. Insurers and auditors ask about the two separately, so one does not count as the other [8].
Conclusion
The frameworks that name a penetration test are fewer than the frameworks that consume the evidence. Only PCI DSS and the FTC Safeguards Rule write the requirement down, but a SOC 2 auditor, a HIPAA risk analysis, an ISO 27001 review and a cyber insurance renewal all end at the same artifact: a dated report that names what was tested and proves the fixes held. Run the 3-Question Evidence Check on the report you already have, and if it fails any of the three, the next test is not an expense. It is the document your next deadline is written against. You can request a scoping call on the Clone Systems contact page and get the test matched to the frameworks you actually face.
References
[1] IBM X-Force. "2025 Cost of a Data Breach Report: Navigating the AI rush without sidelining security," 2025. https://www.ibm.com/think/x-force/2025-cost-of-a-data-breach-navigating-ai [2] FTC. "The Safeguards Rule: What Your Business Needs to Know," ftc.gov. https://www.ftc.gov/business-guidance/resources/ftc-safeguards-rule-what-your-business-needs-know [3] eCFR. 16 CFR Part 314, Safeguards Standards for Financial Institutions. https://www.ecfr.gov/current/title-16/chapter-I/subchapter-C/part-314 [4] Medcurity. "HIPAA Security Rule Delayed to 2027: What It Means for Covered Entities." https://medcurity.com/hipaa-security-rule-delayed-2027 [5] Hightable. "ISO 27001 A.8.8: Management of Technical Vulnerabilities." https://hightable.io/iso-27001-annex-a-8-8-management-of-technical-vulnerabilities [6] Stealthnet. "Cyber Insurance and Penetration Testing: What Insurers Expect." https://stealthnet.ai/cyber-insurance-penetration-testing [7] Clone Systems. "Penetration Testing for SOC 2." https://www.clone-systems.com/soc-2-penetration-test [8] Clone Systems. "Penetration Testing for Cyber Insurance." https://www.clone-systems.com/cyber-insurance-penetration-test [9] Clone Systems. "PCI DSS Requirement 11.4 Penetration Testing Explained." https://www.clone-systems.com/pci-dss-penetration-test-requirement-11-4 [10] Clone Systems. "How to Choose a Penetration Testing Vendor." https://www.clone-systems.com/blog/how-to-choose-a-penetration-testing-vendor [11] Clone Systems. "How Often Should You Penetration Test? Frequency, Scans, and the Retest Nobody Books." https://www.clone-systems.com/blog/how-often-should-you-penetration-test-frequency-scans-and-the-retest-nobody-books
