Skip to content
Services
PCI DSSISO 27001SOC 2GDPRPenetration TestingNIST CSFvCISODORA
Company
Case StudiesPartnersTrust CenterBlogGlossaryFAQFree AssessmentContact
Industries
FintechHealthcarePaymentsSaaS
AboutResourcesVeedar

What v4.0.1 changed and why annual box-ticking is over

On 31 March 2025, fifty-one requirements that had been optional became mandatory. If your last PCI DSS assessment was under v3.2.1, or if you did the minimum to clear that deadline and haven't looked since, your next assessment will be a different experience from your last one.

This is a guide to what actually changed, which requirements most often catch scaling payment companies, and what "continuous compliance" means in practice rather than as a slogan.

What is PCI DSS v4.0.1, and how is it different from v4.0?

v4.0.1 is a limited-scope clarification of v4.0, published in June 2024. It corrects typos, clarifies wording, and adjusts applicability notes. It does not add requirements.

That distinction matters mainly for one reason: people hear "new version" and assume a new project. The substantive change was v4.0, released in March 2022. What made 2025 the pivotal year was not a new document — it was a deadline inside the old one.

What are "future-dated" requirements, and why did 31 March 2025 matter?

v4.0 introduced 64 new requirements. Fifty-one of them were marked as best practice until 31 March 2025, giving organisations a three-year runway. On that date the runway ended. They are now simply requirements, and an assessor must consider them fully during an assessment.

Two consequences that scaling teams underestimate:

If you are assessed against v4.x, all of it applies. There is no partial adoption and no extension. The transition period was the extension.

If this is your first assessment, there is no "before" to be grandfathered into. Everything is in force on day one — which, counter-intuitively, is often easier than retrofitting, because you are designing controls once rather than twice.

The six changes that catch scaling payment companies most often

Multi-factor authentication across the entire cardholder data environment. MFA is required for all access into the cardholder data environment (CDE) — every account, not only administrators, and not only remote sessions. The standard also specifies what the MFA system must do: at least two different factor types, resistance to replay, and no bypass by any user, including administrators, except through a documented exception authorised by management for a limited period. Service accounts, contractor access and break-glass paths are where most gaps live.

Payment page script management. Every script loaded and executed in the consumer's browser on a payment page must be authorised, integrity-assured, and listed in an inventory with a written business or technical justification for why it is there. Separately, a change- and tamper-detection mechanism must alert personnel to unauthorised modification of the security-impacting HTTP headers and script contents of payment pages as received by the consumer browser — at least weekly, or at a frequency you have justified through a targeted risk analysis.

This pair is the Council's answer to digital skimming, and for most teams it means new tooling — subresource integrity, a content security policy, or a script-management platform — rather than a policy edit. If you embed a provider's payment form in an iframe, the scripts inside that iframe are the provider's responsibility; the scripts on your surrounding page are yours.

Targeted risk analysis. Several requirements now let you set your own frequency instead of following a fixed schedule — but the flexibility is conditional. You must document a formal analysis identifying the assets being protected, the threats, and the likelihood or impact of the threat being realised. No analysis, no flexibility: you fall back to the prescribed frequency.

Authenticated internal vulnerability scanning. Internal scans must now be performed with credentials and sufficient privileges. Systems that cannot accept credentials must be documented as such. There is also a defined expectation for handling vulnerabilities that aren't ranked high or critical — they are addressed on a timeline your risk analysis justifies, not left indefinitely.

Automated log review. A person reading logs each morning no longer satisfies the requirement on its own. Mechanisms are expected to do the reviewing, with people responding to what surfaces.

Twelve-character minimum passwords for in-scope accounts, alongside the broader authentication changes.

The customised approach: useful, but not on a first audit

v4.x introduced a genuinely new structural option. Alongside the defined approach — meet the requirement as written — you can now meet a requirement's stated objective with a control of your own design, documented in a controls matrix and supported by a targeted risk analysis.

For teams with modern architectures that don't map cleanly onto assumptions written for 2006-era data centres, this is a real improvement.

Three caveats before you reach for it. It is more work to evidence, not less. It cannot be documented in a self-assessment questionnaire, so it requires a Qualified Security Assessor (QSA). And on a first assessment, it adds a novel argument to a process where you have no track record yet. Use it when a specific requirement is genuinely impossible for you as written — not as a shortcut.

Why annual box-ticking is over

Here is the deeper shift, and it is the thing worth telling your engineering team.

Almost every change above describes a control that has to be operating, not a control that has to have existed once. Monitoring that runs continuously. Tamper detection that alerts weekly or better. Scanning with credentials on a schedule. Risk analyses that are reviewed and re-justified. MFA enforced on every account, all year, with exceptions that expire.

Under that model, a folder of screenshots from last spring is not compliance. It is a snapshot of a moment, and the assessment is asking about the year.

Service providers carry this further still: scope must be validated at least once every six months and after significant change, rather than annually.

The practical consequence is that the old rhythm — ignore it for ten months, panic for two — is now the expensive way to do this. Not because anyone is being punished for it, but because reconstructing a year of evidence after the fact costs more engineering time than producing it continuously ever would.

What to do about it, in order

Confirm which requirements apply to you. Your validation path — which self-assessment questionnaire, or a full Report on Compliance — determines the set. Ask your acquirer in writing rather than assuming; this single email removes the most common source of wasted effort we see.

Get scope right before you build controls. Every system removed from the CDE is a system you never have to enrol in MFA, scan, monitor, evidence or explain. Tokenisation, hosted payment fields and tested segmentation do more for your cost and timeline than any efficiency you can find later.

Run a gap assessment against the future-dated set specifically. If your programme was designed under v3.2.1, this is where the surprises are. Produce a gap register with an owner and a duration per line.

Design the evidence at the same time as the control. For each new control, decide now where the record lives and how you will produce twelve months of it. The teams that find their second assessment cheap are the ones that answered this question before their first.

Write down your targeted risk analyses. They are the key that unlocks several of the flexible frequencies, and they take an afternoon each. Skipping them quietly removes options you would otherwise have.

Frequently asked questions

Is v4.0.1 a new standard I need to re-certify against?

No. It is a clarification of v4.0. If you are already assessed against v4.x, you are in the right document; the change that requires action was the 31 March 2025 deadline for the future-dated requirements, not the point release.

What happens if we haven't implemented the future-dated requirements?

They are assessed as requirements. Gaps become findings, and findings must be remediated before an Attestation of Compliance (AoC) is issued. There is no extension mechanism.

Do the future-dated requirements apply to small merchants using a self-assessment questionnaire?

It depends on which SAQ applies to you — each one covers a different subset. SAQ-A merchants using a fully hosted payment page have a much smaller set than an SAQ-D merchant. This is exactly why the validation path question is worth settling first.

Does MFA really apply to service accounts?

The requirement covers all access into the CDE. Application and system accounts are treated differently from user accounts in several places in the standard, so the honest answer is that this needs a conversation with your assessor about your specific architecture — but assuming service accounts are automatically out of scope is the assumption that produces findings.

How long does it take to close this gap?

For a scaling payment company doing a first full assessment, three to six months end to end is the realistic benchmark. Scope dominates that number more than any other variable.

Still have additional questions about your compliance readiness? Get a free assessment

Comments

Be the first to comment.

Leave a comment