
How often should an online casino run a vulnerability assessment?
Quarterly is the compliance floor, not the answer. The cadence framework for online casinos — regulatory minimums, the AnySec baseline, and the five risk-based triggers that should move a scan off-cycle.
The short answer
Quarterly is the realistic floor for an online casino's vulnerability assessment cadence — it's the minimum PCI DSS 11.3.2 sets for external scanning if you process cards, and the baseline AnySec recommends for any regulated, internet-facing platform. Faster-moving operators, or platforms in a licensing jurisdiction with an explicit scan mandate, should run monthly. But calendar cadence only answers half the question: a new feature launch, an infrastructure or vendor change, an M&A event, or a security incident should each trigger an assessment on its own, independent of where you are in the quarterly cycle.
Who this decision is for
This is written for the person who has already decided to run a Vulnerability Assessment — not choosing between VA and a pentest, which is a separate decision covered in vulnerability assessment vs penetration testing for iGaming — and now has to set or defend a recurring cadence. That's usually a CISO or Head of Security setting a scan schedule for the first time, defending an existing schedule to a board or auditor, or deciding whether a one-off engagement should become a recurring one.
It's not for you if you're still defining what an assessment covers. Scoping an external vulnerability assessment for a casino covers what gets scanned and where an external-only scope's evidence runs out; this post assumes that scope question is settled and answers a different one — how often to run it.
The calendar floor: what regulators and standards actually require
Two concrete, checkable minimums, plus the AnySec baseline:
| Requirement | Cadence | What it covers |
|---|---|---|
| PCI DSS 11.3.2 | Quarterly | External-facing scan, if you process card payments — must be performed by a PCI SSC-listed ASV if the deliverable needs to satisfy that specific requirement |
| Pennsylvania 58 Pa. Code § 809a.6 (Interactive Gaming Platform Requirements) | Quarterly, plus after any Board-approved change or modification | Internal and external network vulnerability scans, with quarterly verification submitted to the Board |
| AnySec recommended baseline | Quarterly minimum, monthly for fast-changing or regulated platforms | Full external-plus-internal estate, manually validated |
The Pennsylvania rule is worth calling out specifically because it does something PCI DSS 11.3.2 doesn't: it builds a change-triggered clause directly into the regulatory cadence, not just a calendar one. That's the same logic a risk-based cadence should follow even where the letter of the law only asks for quarterly.
Licensing requirements are not uniform across jurisdictions — treat any specific number here as one concrete data point, not a substitute for checking what your own regulator requires.
Beyond the calendar: five risk-based triggers
A calendar cadence catches drift over time. It doesn't catch the moment your risk actually changed. Five events should move a scan off-cycle regardless of where you are in the quarter:
- A new feature or platform launch. New attack surface — a new cashier flow, a new API, a new affiliate integration — is riskiest in the window before it's been battle-tested, not three months later when the next scheduled scan happens to reach it.
- An infrastructure or vendor change. A new hosting provider, a new payment processor, a new third-party analytics or BI tool — each one is a new set of assumptions about what's exposed, and the kind of internet-facing internal tool CVE-2026-72898 in Metabase showed can carry production database credentials without anyone flagging it as in-scope.
- M&A, or expansion into a new licensing jurisdiction. Acquired platforms and newly-licensed markets both mean estate you haven't assessed yet, on someone else's patch cadence.
- Post-incident remediation verification. A scan that only confirms the specific finding was fixed misses whatever else changed during incident response; a full re-scan verifies the fix actually held.
- A CVE lands in a component you run. Not every disclosure warrants an unscheduled scan, but one in a component you can confirm is deployed — especially one heading toward a KEV listing — is exactly the kind of event a purely calendar-based cadence structurally can't react to in time.
What a mature cadence looks like over time
Cadence isn't static — it should track how the platform matures:
- Starting cold: a single external VA to establish a baseline before committing to a recurring schedule.
- Established, standard risk: quarterly, matching the PCI DSS floor and giving four checkpoints a year to catch drift.
- Fast-changing or regulated (casinos, fintech, crypto): monthly, because the estate moves faster than a quarterly cycle can track.
- High-risk platforms generally: recurring VA on whichever of the above cadences fits, plus one annual penetration test — VA catches breadth and drift, the pentest proves exploitability against the flows that actually matter.
What you get each cycle
A Vulnerability Assessment engagement runs 3–5 business days per cycle, with a quarterly or monthly recurring option, and produces a prioritized, de-duplicated findings list ranked by real exposure rather than raw CVSS, concrete remediation guidance per finding, and an audit-ready summary formatted for a regulator, payment partner, or internal audit trail. Every automated hit is manually verified before it reaches the report — findings volume doesn't compound just because the cadence does.
Limitations
- This is a cadence framework, not a scoping decision. It doesn't tell you whether to run VA or a pentest — that's already covered — or what an external assessment specifically checks — also already covered.
- Regulatory minimums are jurisdiction-specific. The two cited here (PCI DSS, Pennsylvania) are concrete examples, not a survey of every licensing regime; confirm your own regulator's requirement directly.
- More frequent scanning without remediation capacity doesn't add security — it adds backlog. A monthly cadence only helps if findings get triaged and fixed between cycles; otherwise it just produces more unread reports faster.
- Cadence doesn't substitute for depth. No scan frequency, however tight, replaces the goal-driven testing a penetration test does against your cashier, wallet, or auth flows specifically.
Decision framework and next step
Quarterly is the floor. Move to monthly if you ship fast or sit in a regulated sector. Layer the five risk-based triggers on top of whichever calendar cadence you land on, regardless of when the next scheduled scan falls. Add one annual penetration test if the platform is high-risk. If you're not sure which of those applies to your platform, tell us your estate and current schedule and we'll help you set — or defend — the right cadence.
Sources and review
Pennsylvania's quarterly-plus-change-triggered scan requirement is cited to 58 Pa. Code § 809a.6, System requirements, the Commonwealth's own published code. The PCI DSS 11.3.2 quarterly external-scan requirement is cited to the PCI Security Standards Council's own Approved Scanning Vendor program page, the same source used in the external-scope post linked above. Author: AnySec Engineering. Published 2026-08-19.
Related reading
- Vulnerability assessment vs penetration testing — the breadth-vs-depth engagement-type decision, before cadence is the question.
- Scoping an external vulnerability assessment for a casino — what an external scope actually scans and where its evidence runs out.
- CVE-2026-72898: admin takeover via Metabase — a real example of the kind of vendor/infrastructure change that should trigger an off-cycle scan.
Rather not learn this in production.
Talk to the engineers behind these write-ups — thirty minutes, no sales script, a straight read on where you stand.
Book a call
